The prototype was destroyed during the 9th test flight after developing an engine fire on the #3 engine. All further work was canceled, and the increase in speed was not deemed to be worth the added complexity and weight of the watercooling systems.
I think, more to the point, the added speed wasn't considered sufficient enough to interrupt production to get the B-38 in production. Note the leading edge radiators between the engine nacelles. This modification probably made it less viable than if they used annualr radiators on a quick engine change module, as the Germans did.
The V-1710s were considered for the B-40 escort bomber gunship. The extra speed woul be handy to overcome the weight deficit that the YB-40 had over B-17s once the latter had dropped their bombs. In any case, the B-40 concept was soon dropped.
In the case of the XB-39, the nacelle was developed as a QEC by Fisher to replace the R-3350. They were tested on the Douglas XB-19, but with experimental turbos. For the XB-39 the turbos weren't fitted, altitude rated V-3420s were used instead. If the XB-39 used twin B-series turbos, as did the Wright R-3350 powered B-29s, performance may have been better. The reason for not adopting the B-39 was that it too would interrupt production. Not sure by how much, since the QEC was developed as a direct replacement for the R-3350 module. I would imagine that they could have been retrofitted, if need be, elsewhere.
Perhaps a more problematic issue is the priority level the Army gave the V-3420. Over its development life it was high priority, low priority, cancelled or prioritised for other programs. For instance, when Schweinfurt happened it suddenly dawned on the USAAF that they may actually need a long range escort - thus the P-75 program was born. And it took priority for V-3420s over the B-39 program, and since Fisher was developing the P-75 it delayed their work on the B-39 conversion.
Originally the B-15 was to be powered by V-1710s. It was thought in the early '30s that air cooled engines did not have the cooling capacity to deal with the boost levels required for high altitude flight, which were to be provided by turbocharging. But the V-1710s were not ready in time, and the XB-15 flew with R-1830s. Some say (like Wiki) that it was originally intended to use V-3420s, but these were not on the drawing board at the time. Dan Whitney's
Vees for Victory states that the V-1710 was selected for both the XB-15 and Martin XB-16.
Also in the '30s it was felt that engines buried in the wings of aircraft would produce less drag than those in nacelles. Thus the Lycoming O-1230 and Continental O-1430 (later to become the IV-1430) horizontally opposed engines were developed to be buried in wings. A similar requirement led to the Wright R-2120 Tornado.