The lighter Ash-82 replaced the lighter M-105; the heavier R-2800 would replace the heavier V-1710. The intercoolers can go to to the place previously occupied by the cooler radiators.
We can also take a look at the Spitfire, that 'swallowed' far heavier engine (Griffon) without too much trouble.
I've already covered the installation of the intercoolers, so they should balance the rest of the powerplant decently.
As far as the prop goes, P-47 was using the 12,2 ft prop, 4 bladed, early on. NAA can be smart and use the system from the P-47 (extend able gear strut).
The M-105 was actually fairly close to the V-1710 in weight. According to some sources 1268lbs. With the radials there is some shift (small) in the engines center of gravity rearward in relation to the aircraft's center of gravity. The La-5 is about 14cm shorter than a Lagg-3, not a lot but every bit helps. The La-5 was nose heavy enough that armament was limited and we don't have good details on other equipment sights or ballast. P-51 intercooler and radiator were filled with a water/glycol mixture which weighs a lot more than air

Sticking the intercoolers under/behind the pilot gives you a P-47 without the turbo. You don't need the exhaust ducts but you need ducts to carry the intake air from the first stage of the supercharger to the belly mounted intercooler/s and then ducts to bring the intake air back to the engine carburetor/second stage. Intake air inlet needs to be near the 1st stage supercharger section. Cooling air for the intercooler can be brought in from under the engine (F6F or P-47) or from under belly scope but the intercoolers for the R-2800 do not weigh anything like the radiators on the Merlin unless you design and build a air/liquid intercooler for the R-2800 instead of the air/air system that they did use.
You could use 12'2" props and extending landing gear, But every change from the Allison powered Mustang is more engineering time and more testing.
Griffon required much larger radiators behind the CG than the Merlin did which helped balance things, plus ballast.
The R-2800 powered P-51 would be far less draggy than F4U, far less weight, with more internal fuel (assuming the disposition of the fixed fuel tanks as it was in 1944), it's able to carry 2 drop tanks from day one (unlike the P-47). Granted that it would be using more fuel than the real Merlin P-51, but also far less than the P-47/F4U.
I am not sure where the far less drag is coming from. Yes you have a smaller, lighter wing but moving the intercoolers from behind the engine to the rear/belly of the fuselage may not be a low drag option. We know that a radial P-36 had 22% more drag than P-40 so even if the bigger diameter (and greater appetite for air) R-2800 installation only has 12-15% more drag than the Merlin that will mean a 12-15% higher fuel burn even if you can keep the R-2800 from having to operate at a rich mixture setting.
A P-51D clean can do 335mph at 25,000ft burning 59 gallons an hour. A Corsair at 21,500 ft and doing 282mph is burning 61.2gallons and hour and at 338mph at 21,500ft is burning 127 gallons an hour. It has to shift into rich mixture to hit this speed.
A P-47 at 25,000ft clean burns 95 GPH to do 300mph and 145GPH to do 337mph.
Of course, but so is the 2-stage Merlin for the P-51.
True, neither one should be a consideration for what engine to pick in 1940.
2. 1-stage V-1710 for Mossie, why? Because the USA is not building Merlins, so the RAF/MAP need to do some reshufling of those. I was thinking more about the 8.80 SC ratio engines, not the really low level ones.
3. NF job is obvioulsy not for the 1-stage V-1710s. The strike fighter can use the lower weight and lower drag of the inline to make good for the loss of the ~10% of the power, and can also use the WEP (1590 HP at 4500 ft, no ram) for the, well, war emergency

4. Not really so, with the 8.80 SC ratio engines. That should do up to 15000 ft.
The Allisons are are bit late to the game. There are two basic models using 8.80 gears that need to concern us. The F3R (-39) model which was good for 1150hp take-off and 1150hp at 11,700ft, While not in production in 1940 we may assume it was in the planning stages. WER is not approved in US service until Dec 1942 although the British were using high than book pressure much earlier. These engines do not show up in production until the summer of 1941 and are eventually cleared for 1490hp at 4300ft.
The Merlin XX which was coming into British service in the summer of 1940 and was the Engine that Packard signed to produce was good for 1280-1300hp for take-off and 1240hp at 11,500ft low blower and 1120hp 18,500ft high blower at 12lbs boost. If the British use the higher boost limits for the XX they can get 1480hp at 6,000ft low blower and 1480hp at 12,500ft high blower. The later F4R Allison engine used the P-40 K was strengthened and while rated at 1325hp for take-off and 1150 at 12,000ft it was rated for 1580hp WER at 2500ft. While the beginnings of this version go back a ways it did not enter production until late 1941 and the first P-40K with one does not roll out the factory door until May of 1942.
Even with 8.80 gears too much power is being given up at all but the lowest altitudes (under 5000ft) to make this a good choice. Coupled with the 10% or so loss of power for take-off (unless over boosting is used) and it really looks less than attractive.
Obviously the British would not re-engine all of their engines, but the designs that would still been serving them good in decent numbers. Using the US engine would free up their engine production for the other designs where such a change would prove troublesome, either from the production or performance standpoint.
Substituting R-2600s for Hercules engines might be simpler. It at least solves a big part of the power problem for the 4 engine bombers and thus frees up Merlins for other uses.
You mean with Packard-built R-2600s? IIRC Beufighter was tried with R-2600 historically.
I believe there were a few other airframes that got R-2600s on an experimental basis. There was a plan to build Short Stirlings in Canada powered by R-2600s, with two airframes converted. edit> or 4 airframes with 2 flown?<
The R-2600 maybe a viable replacement for the Hercules.
Edit> Later versions of the Merlin XX include the Merlin 21 used in early Mosquitoes, pretty much the same ratings as the XX, The Merlin 23 with 1390hp available for take-off and 1435hp at 11,000ft using 16lbs boost, used in the Mosquito Mosquito I, II, IV, VI, XII and XIII and the Merlin 25 with 1610hp for take off and 1510hp at 9,250ft using 18bs boost used in the Mosquito VI and XIX. Packard versions sometimes varied by 20-30hp but Packard engines include the Merlin 31 and 33 ( a Merlin 23) used in Canadian and Australian Mosquitos. The Merlin 38 used in Lancaster Is and IIs (1400hp take off) and the Merlin 224 225 also used in Lancasters Mosquitoes respectively with 1635hp for take-off and 1680hp at 2500ft in low blower at 18lbs. None of these are two stage engines.
In 1940 these higher powered versions were in the future and needed 100/130 fuel and engine modifications to reach 18lbs of boost. 12-14lbs boost was achievable with with British 1940 100 octane fuel.
Perhaps the V-1710 could have been developed to reach these power levels. A two speed supercharger drive would have helped take-off power a lot. It would have been worth an easy 100 hp for the take-off ratings.