They are dive
flaps and not dive
brakes. They do not appreciably slow the aircraft. Maybe I need to try it again, but in a dive from 20K, in AH, and in AH II, they do not create a hands free 4G pullout. They never have, in my experience. Buffeting should begin just before .64 MACH, and provided the dive flaps are extended by .67 MACH, the buffeting should ameliorate or at least lessen, and the plane should start to recover once you get in thicker air. In fact, most pilots who knew what they were doing said if you stuck with the plane and flew it instead of panicking said that once it got into thicker air, you could pull out without the flaps. The dive flaps themselves do not actually affect the shock wave of compression, nor how or where it forms. What they do is alter the center of lift on the outer wing, so that the plane can regain a nose up attitude. Note that the dive flaps are on the outboard wings, but the thick wing section on a P-38, where compression happens, is the center wing, between the fuselage booms and the center nacelle. THAT is where compression occurs, as stated later in this post.
Witness Robin Olds' story
http://www.youtube.com/watch?v=ITRLk9b9AcY of shooting down two Me 109's and then diving to attack two on the tail of a P-51 and entering a severe compression dive. He pulled out (it blew out a window, which was common, the MACH shock wave that causes compression forms where the wing fillet and the cockpit nacelle join at the window) leveled off, and proceeded to shoot down another M3 109 that attacked him at treetop level. Notice there is no mention of dive flaps.
Note, Dogfights is wrong about several things here, including the "approaches the speed of sound" (the P-38 would not approach the speed of sound, it went into compression at around 490 MPH), and the visual of compression and what happens during compression, it does not cause elevator flutter or lock the control surfaces, it renders them ineffective because it shifts the center of lift, forcing the nose down. Now, Robin Olds was wrong, too, he thought pulling G's blew out his window, but the cause was compression and the shockwave. If you read Warren Bodie's "The Lockheed P-38 Lightning" you'll find that the radius of the center wing fillet where it joins the center nacelle, and the fit up around the roll up window on the side of the cockpit, are both the most critical components in how and when compression effects the P-38. In fact, had that radius been increased and extended again, and the cockpit design altered slightly, it might have improved the compression issue further, and raised the threshold again.
Another thing I found seriously wrong with the P-38, although it is a somewhat minor flaw, is that the wheel brakes just plain suck. The proper run up procedure has the pilot hold the plane at the end of the runway and build 54" of boost, while the plane remains still. This was to prevent the problem of engines stalling or cutting out as the throttles were advanced no take off. I've seen it done, repeatedly, and all of the current pilots flying restored P-38's that I've seen practice this religiously. I watched Steve Hinton do this to Glacier Girl dozens of times, the plane sits still and shudders while the engines howl, then the brakes are released and it rockets down the runway.