Thanks for the link, McPherson. Reading that old thread (20 years old!) makes one realize just how tired much of the P-38 improvement arguments have become, especially the idea of "dropping" Merlins into it. Throughout my efforts here I have learned a lot about the Allison and how they compare to the Merlin and the truth is, they are often unfairly underrated verses the Rolls-Royce, which has attained some sort of
wundermotor status. As
@tomo pauk has pointed out elsewhere, the Allison's primary failing was lack of proper two-stage supercharging. When a second induction charge stage was added it was either done by an auxiliary supercharger driven by driveshaft (P-63) or by a Turbo (P-38). The first method was adequate, but compared unfavorably to the Merlin's fully integrated and compact packaging. The latter, of course, we know well: installation was complicated and prone to pressure loss in the ducts and more susceptible to damage (larger area = more likely to be shot). All the other problems with the Allison were solved or balanced by problems with the Merlin--which are almost always overlooked in these types of discussions.
Take the oft-discussed fuel issues and lead-separation with the turbo-charged Allison. These were related more to the intercooler installation (which, let's remember, was designed with the MTO and PTO in mind where low-altitude, hot day, operation was normal) than to the engine itself and once better fuel mixing and additives were used beginning Spring '44 were all-but non-existent. Conversely, these same "better" fuels tended to play havoc with the Merlin valves and resulted in an increase in engine failures and a reduction in MTBO. Let's also not forget that the Merlins tended to be a bit more temperamental, especially the British built variants, thanks to their tight engineering tolerances where the Allison was a bit more forgiving. Plus, there is the Power:Weight. The Allisons were generally (depending on the installation) lighter per HP, certainly an advantage in an airplane.
All of that aside, people also tend to ignore that putting even the best Merlins in a P-38 wouldn't fix the Compressibility problems, which was the airplane's biggest limitation. At 30,000', the Lightning could approach its Critical Mach
even in level flight. All it would take is a short, shallow, dive and it's there. I think applying a proper fix to that issue is needed before any other major changes even begin to make sense.
The Twin Mustang wasn't really designed as an interim fix for anything but to fill a very specific need which no other aircraft could do: high-altitude, very-long range, escort. It was designed to escort B-29s and B-36s round-trip on their missions and to this day still holds the Piston-engine un-refueled Hawai'i to New York (5051 mile) record at 14 hr 32 min. It was only later that they realized it wasn't really needed in that role, so they found other niches for it to fill in an interim basis, specifically as a Night-Fighter or Radar equipped all-weather interceptor. They had some success in both roles during the early stages of the Korean War until purpose built jet aircraft became available (e.g. F-94).