My $.02 on an early P-38: Whenever I see posts requesting early 38 models in this forum, the author invariably asks for an F or occasionally an H. My feeling is that the G-10 would be the most appropriate variant to model.
More Gs were made than Fs and Hs put together. 1282 of the P-38G were made vs. 377 F and 601 H models. Of those 1282, 548 were the G-10 sub-variant (more than double the production of any of the other 5 sub-variants of the G). The G-10 added provisions for mounting twelve 4.5" rockets in four three-tube "bazooka" type launchers and also recieved winterization equipment (not sexy but very nice to have in the Aleutians and the ETO).
P-38Gs began rolling off assembly lines in November 1941 and were in service at the squadron level by spring 1942, so they wouldn't be out of place in any early war USAAF scenario.
Of course I'd be thrilled if HTC saw fit to model an F, a G, an H, and a J. But, judging from the fact that we're nearly 3 years from beta and still only have 1 P-38, I'd say the chances are good that we'll only get one more. If that is indeed the case, then it makes sense to me to model one that is different enough in performance to the P-38L to be noticable.
A P-38J would be the easiest for HTC to make. No alteration of the 3D model would be necessary (assuming it's not a -5), just paint it green, remove the dive flap light from the cockpit, tweak the high speed roll rate, and voila! P-38J-10, -15, or -20. Other than the roll rate though, the J performs so much like the L that it makes me wonder if its even worth the effort to make those minor mods.
A pre-J model on the other hand, would require re-worked engine cowls, radiator scoops, and canopy. More work for HTC, but the result would be a Lightning that looks AND performs quite a bit differently from the current one. 400mph vs. 414mph and 8.5 minutes to 20k vs. 7 minutes (G vs. J/L).
OK, I guess I've rambled on long enough. In summary, my stance is that when/if we get another P-38, it should be the P-38G-10-LO because it was the most produced of the early variants, was in service in 1942, was the first equipped to operate in any climate, with the 12 rocket option it would retain an acceptable jabo capability (unlike the P-47D-11 and F4U-1), and its performance is different enough from the L to justify the work required to model it.
P.S. What Mitsu said.
