I don't know man, in every A&A game I've ever played, but especially in AA50, Japanese players consistently crush towards Moscow. Has nothing to do with how many spaces are in China (not enough to stall anything, at least in 41 for AA50, and again in 1942.2) or whether or not India has a starting factory (an even stronger target in 1942.2 than the same territory with no factory in AA50, since UK doesn't have the money to stack it early enough in 42.2), either way I still see Japan do the same thing it does in all A&A games since Classic, e.g. center crush. They might not necessarily crush Moscow along the northern route, they might push middle or south, but always with the same ultimate goal of linking with Germany at the center. Japan has every strategic incentive to do this, and virtually nothing to discourage them from doing so.
A NAP could conceivably help delay this (Moscow crush), but it can't prevent it entirely unless you change the victory conditions. Only if Axis or Japan specifically, has another way to win the game, can you alter the center crush dynamic. This was attempted in G40.2 (the 6 VC Pacific win), with somewhat mixed results. First the NAP in G40 sec edition tends to fall apart, everyone I know immediate declares war on Japan with the Soviet Union, because there's no reason not to (no strong penalty for the Russians, or a boon for Japan to make them second guess it). The only thing that holds it back from a practical standpoint is the way Amur attaches to Mongolia activation. But if the 6 VC sudden death win is out, for any reason, then Japan will immediately Center Crush (many people do this regardless and focus on propping up the European Axis as the safest play long term). I believe this is because of the way the Allied Victory conditions are written in G40.2. In G40 sec ed Axis determine how long the game goes on, because there is no sudden death win for Allies comparable to the sudden death win for Axis. It basically devolves to eithee Axis sudden death, or Axis concession, but there is no way for Allies to win outright, short of capturing all 3 Axis capitals. Since 3 capitals is unreasonable from a game length standpoint (requires a masochist Axis player with supreme patience and a total glutton for punishment hehe), this means Japan has every incentive to crush center, because they still have a chance to influence the war in Europe far more easily than say Germany/Italy can influence the war in the Pacific. And it's possible to win on the Europe map alone for Axis if they take 8 VCS there, regardless of what is happening in the Pacific.
In 1942.2 there are no such conditions to seperate VCs by theater, and no real way for Japan to do anything other than pressure Moscow as an endgame strategy. They cannot viably contest North America. So my concern is that any NAP you might implement, absent alternative victory options, would just totally hamstring Axis.
I'm not sure, it's something I have given a lot of thought to, but still haven't found an elegant solution for. The way we played with a NAP, was essentially exactly the way you outlined. But the problem is that it becomes very confusing, with armies fighting each other in occupied territories while ostensibly still neutral towards each other. This is where I think it becomes problematic to present the NAP as something that can never be broken, or a strict rule. I favor a cost/benefit structure, where you give players an economic incentive to preserve neutrality between Japan and Russia, and an economic disincentive to break with neutrality. The trick is that these costs/benefits have to outweigh or at least roughly match, the economic or strategic advantage of declaring war. I think it needs to be fairly high in terms of IPC (as a one time bonus) or else consistently decent over time (a bonus that applies every round the NAP is in effect.) But still retaining the option to break the Pact at some point if desried. Since that happened in 45 after all.
Right now I see Japan do the "Burma push" in 1942.2, or the blast through Szech, or both at the same time. But Szechwan is the shortest distance from the coast to Moscow on this board. A strict NAP (an unbreakable NAP) shuts this route down, leaving only the south via India. Once taken, it's very hard for Russia to do much of anything along this front, they can't liberate India after Japan stacks it in my experience. Then Japan just launches fighters to Europe or mobs Africa or hops into Europe from the Med (since they can't get there through Russia.) With the NAP played like this USA would probably want to go 100% Pacific, but I don't know that they have the resources to pull off a KJF if Russia can't support it directly, in the first round. Seems a stretch don't you think?
I'd be interested in hearing more ideas on this subject as well, since I really enjoy the 1942.2 board.
Also it's probably worth mentioning sz 16 here as well. Everyone I have discussed the subject with seems to agree that sz 1 closed is better for game "balance," though open tends to be more fun for game "play." The problem with leaving sz 16 open for warships and transports, is that it allows G to center crush all by itself, before anything of consequence happens elsewhere on the board for all the other players playing. This basically takes the whole game down to a few early battles on the Eastern Front, with advantage going to G unless Russia rolls miraculous defenses. I prefer sz 16 restricted myself, as I think closing the bosphorus is critical for Russia to have any chance of surviving early on.
Great replies so far. I'm eager to hear more houghts on these subjects, since 1942.2 is my go to starter board for training new players.