I found a BuAER document, dated January 25(?), 1941, that discusses an upgrade to the F2A-3 to make it more competitive. Upgrades to include:
1. wing modification to incorporate a full span NACA slotted flap and attached aileron slot lip ailerons.
2. installation of an R-2600-12 engine to a fuselage lengthened by 18 inches. wing to remain standard F2A. However entire airframe to be stregnhened accordingly.
3. New F2A wing to be folding with accomodation of two .5 inch HMGs in each one.
A maximum speed estimate of 403 mph.
Modifications expected to produce a gross weight of 8,185 lbs.
Attached was a copy of a BuAer letter of Feb 17, 1941 rejecting the proposal as unattractive. Thank GOD!
Stall speed calculated to be 80+ mph, range 975 miles with extension to 1085, below the desired 1,500.
Yes, this seems like a pipedream, though I suppose that's not inconsistent with some of Brewster's design projections and promotion. Best-case, they were heavily exaggerating how much it actually had in common with the F2A, was basically a new aircraft, and might have actually been better suited to Brewster's production facilities (and quirks) than the F3A Corsair production turned out to be. (even so it wouldn't have managed those performance specs and would've likely been plagued by the same problems other R-2600 platforms had with various corrosion issues, carburetor mismatches, etc)
I suppose it also would've been a better use of resources than the SB2A turned out to be, but that's not saying much either. (aside from hypothetically allocating a similar number of R-2600 engines)
Their efforts with the XA-32 weren't really encouraging either, though.
I'd say the only thing they'd really have been suited for was fixing the basic problems with the F2A, changing as little as possible to get something reliable and useful, being as conservative as possible, and working on licensing it out to offset in-house production difficulties. (albeit the latter probably should've been done with the F2A in the first place, rather than expanding internally ... the Navy had already used the Naval Aircraft Factory to supplement Brewster production with the SBN, and while I'm not sure the capacity there would've been suitable for F2A production, licensing the design to some other significant manufacturer might have been viable ... perhaps Curtiss would've been better suited to that than it turned out they were with P-47 production)
I could see a rationalized F2A filling the escort carrier fighter role similar to the F4F, but without folding wings it would have to rely on smaller natural footprint and perhaps adopt slightly shorter, squared-off wings. Beyond that, they needed to focus on reducing weight as much as possible, compromising with a more limited internal fuel capacity, and simplifying some of the components for more streamlined assembly.
I recall the wing being both a pain to assemble in the factory and service in the field, so modifying that might have been the biggest change. Part of that would've been deleting the integral wing tanks, adding appropriate conventional internal structure to fill the structural duty, and reserve more space for newer, self-sealing fuel cells in the wings. (the F2A-3 added a limited amount of self-sealing tanks in the wing leading edge and in a fuselage extension, but beyond that had to rely on the unprotected main tanks, of 8- gallons each iirc, but one of which was normally capped and unused other than ferry flights or very long patrols)
They probably could've deleted the ventral window to simplify things and possibly save some weight, too (or improve structural integrity at similar weight). Deleting that window might also have facilitated some of the wing redesign. (ie not needing to have any significant view below the pilot's feet opens up a lot more engineering flexibility) I'm not totally sure why they included it in the first place, but it seems like the US Navy was interested in visibility below the aircraft as the F2A-3 also featured ventral view ports, albeit much smaller ones than the F2A sported.
The landing gear issue was supposedly solved by reducing the tire pressure (which screams insufficient testing and poor documentation) but skimming off some of the weight in general should have further moderated that problem.
The single-stage R-1830 might have been a better fit for the F2A than the R-1820, especially the lighter single-speed supercharger variant. It wasn't lighter than the R-1820, but the smaller frontal area should have complemented the F2A more than it did the F4F, and I'd argue those engines would've fit the F2A better than they did the various Wildcats and Martlets. (and a few F4F models using single-stage superchargers) However, given the likely engine mount and CoG changes, that's more a change that might have been worthwhile prior to the F2A-2's introduction (ie developed and tested as a follow-on to the F2A-1) and by the time the F2A-3 was in service, it might have just been better not to bother and instead focus on not screwing anything else up. (as it was, they might have managed that had they cancelled all projects except the F2A and just focused on making it work reliably)
Had the F2A-2 (onward) been designed around using engine mounts compatible with either the R-1820 or R-1830 (at least within a certain weight threshold), it would've made production a good deal more flexible as well, more like the Wildcat or P-36/Hawk-75.
Also note the F4F also had landing gear problems, but those were related to the manually cranked mechanism employed and pilots failing to make sure they'd fully locked the gear in the down position before landing. (I don't believe that issue was ever fully solved, or an automatic/powered gear mechanism ever installed, though I now wonder how much weight/complexity the manual system saved and if that might've been a solution to some of the F2A's problems)
I'd also think the F4F would've been a much better platform for the R-2600 given the aircraft's size, structure, and larger wing area. (though the R-2000 might have been more appropriate, possibly moderately better than the up-rated R-1820-56)
Oh, and given the tendency for Navy operations to take place at medium altitudes in most cases, I don't think the 2-stage R-1830 was that critical for the top-gear supercharger performance, though the boost given by the smoother power curves of the 2-speed + neutral as well as presence of an intercooler made some difference and the heavier base weight of the F4F's airframe made the advantages of the lighter single-stage engines less significant.
The R-1820-40 of the F2A-2/3 also had pretty good altitude performance for a single-stage engine, more in the realm of the Merlin XX series and better than single stage R-1830s and the more powerful -56 had a similar altitude power curve, iirc, though better power all around. (critical altitude for WEP with water injection was rather limited, of course)
The FM-2's altitude performance was also worse than the F4F-4 with 2-stage Twin-Wasp, yet it was generally seen as a superior performer. (ie it was less underpowered at most useful combat altitudes)
Alternatively, Brewster also could've just been relegated to lend-lease duty and produced the B.339, 439, etc for export exclusively. As it was, I imagine the Royal Navy would've appreciated Buffalo Mk.I fighters had they retained their naval equipement (and having that historical data point would've been rather interesting to compare/contrast today), but the land-based configured fighters they got weren't suitable for that, primarily due to the deletion of the arrester hook.
It's also worth noting that the Buffalo/B.339 in British Service received self-sealing material (similar to linatex used in British service, or the material used on P-40B/Tomahawk tanks) covering the existing metal tanks. The F2A-2 did not receive that upgrade and the F2A-3 apparently also lacked it (it got more advanced self-sealing tanks in the fuselage and wing leading edge, but the integral wing tanks just got CO2 purge).
With that in mind, it also might have been better had the F2A-2 simply had the self-sealing material added and not the other fuel tank additions or the added weight those cost it. That material worked well against rifle caliber bullets, but failed against .50 cals (or at least US .50 cals ... I'm not 100% sure the lower velocity Italian/IJA 12.7x81mm stuff had enough hydrostatic shock effect to cause fuel tank explosions like .50 BMG). I'd also think it would've been easier to retrofit to the existing F2A-2s, and the lack of having to change the internal structure might have helped quality control. (ie actually refine existing manufacturing rather than introducing new changes)
The US Navy might have found that compromise unacceptable, though, which again points to lend-lease being more appealing.
Also found the transcipt of an interview with Gordon Firebaugh:
Brewster Buffalo pilot Gordon Firebaugh
"GF: Now, while I was with the Chiefs (VF-2) I flew simulated dogfights and whipped F4F's, probably the F4F-3 with no folding wing, but I could also outfly the F2A when in an F4F as well - it all depended on who was in the pilot's seat. I flew the F4F after we got switched around [when the VF-2 NAP's were spread around with other squadrons -jm] in early 1942. I got shot down over Santa Isabel [7 August 1942 during the Guadalcanal operation -jm]...and I've often thought that,
I wish, I'd been better off in a Brewster. I think it would have matched the [Mitsubishi] Zero - the F4F was heavier and didn't have the turning radius. During that fight, I met up with five Zeros, shot down three before I got shot down. I spent a long time in the water, until I got to land and met one of the [Australian] coastwatchers...
GF: Well, remember,
I'm thinking of the F2A-2. We had the F2A-3 for a couple of months,
that was a different aircraft. It had too much fuel. I remember we could fly five hour patrols....
Q: Did you ever get an explanation for the extra tankage in the F2A-3?
GF: They had put in a wet wing - you were able to purge it with CO2 into the main tank, but it meant extra weight. That was maybe the reason we had strut failures - these wheels, the landing gear, landed pretty hard, negative 3 G's. The struts had a tendency to move forward. When you retracted the gear on the next flight, the box strut scraped on the wheel well. You couldn't have that happen, the gear not retracting, so the mechanics would file some off and get closer to the rivets..
Q: And if you did that enough times...
GF: Exactly, you have a gear failure.
I loved the F2A-2, and wasn't as impressed with the -3 and the F4F. Now you know, VF-3 got the first batch and then we got more.
And given this sort of commentary, it seems pretty clear a lot of pilots would've liked a spruced up F2A-2 more than anything else, including simply adding external self-sealing material to the fuel tanks and replacing the telescopic gunsight with a reflector one. (though some might've actually liked the telescopic sight) I'm sure some appreciated the ammo counters on the fuselage machine guns as well.
Even with the F2A-2's fuel capacity, it was probably the longest-legged US fighter in 1940, and the only one close to the A6M-2 ... and still much more heavily built, if still light by US standards. (perhaps more in the vein of BoB era European fighters, though it also had the outstanding roll-rate typical of American fighters)
I'm also unsure about the stability issues mentioned later in this thread related to empty ammo boxes. It might have been specific to the F2A-3, though, and related to the added fuselage extension (between the pilot and engine bay) that included the added fuel tank. If it was also a problem on the British Buffalo, I'm also unsure ... though I imagine that also would've meant adding ballast in examples that installed .303 nose guns. (and the lighter ammo weight of those guns would've shifted CoG a lot less)
The F2A's nose guns also carried up to 500 rounds per gun, so that's a lot of weight of .50 cal ammo, so I could see some validity in problematic COG shift. OTOH, that might have been a sound argument for installing .30 cal guns there instead, particularly given they'd be firing more than twice as fast and been rather effective against Japanese opponents, especially if incendary ammo was used. (the poor rate of fire of synchronized .50 M2 Browning guns made that somewhat appealing in general, at least until self-sealing fuel tanks were introduced: again the sort of external self-sealed tanks used by the Japanese would tend to burst/explode from hydrostatic shock if .50 BMG rounds hit them: the same, incidentally, was the case for the tanks used by German bombers during the BoB and likely would've been substantially more effective than the .303 guns for that reason, even if the effect on structural components wasn't so drastically different or anywhere neat that of 20mm shells) Also, unlike .303 British, the trajectory of .30-06 was much closer to .50 BMG, so harmonizing the guns (and using .30 cal guns to range-in) would've been rather useful.
The F2A-3 itself, with its max fuel load, might have been more useful as a photo recon plane. Otherwise it wasn't all that useful or practical an advantage, including the lack of bombload sufficient to really merit the added range. (in the latter case, had it been capable of carrying as much or more bombs than the Hurricane IIs in Commonwealth service, it would've made a more useful tactical bomber while Hurricanes could fly escort ... but with the roles reversed, the range of the Hurricanes were far too limited for the fuel tankage of the Buffalo to matter at all)