I Consider the G4M2 a waste of resources. The Frances arrives earlier, it's much faster (340 mph) and it has armour. It lacks the range of the Betty, but for long range missions you already have the G3M3 (Nells) which can go up to 26 hexes and it's more or less as good as the Betty.
Moreover the Frances uses the Ha-45 engine, which is used by most mid-war allied fighters, so you should have enough stockpiled when the R&D industries for the Frances become repaired.
I have also concerns about researching the Judy line and the Myojo at the same time. Don't think it's a wise solution. Why invest so much on the Myojo and so little on the D4Y1? The D4Y1 can be researched easily with 4x30 factories and get it by november 42 without much stress. Then you better skip the D4Y2 (which keep on using the Aichi-60 engine) and start R&D the D4Y3 (which, instead, uses the Ha-32 which can be stockpiled up to 500 easily). The latter will be, imho, the key for late war naval engagements. With the engine bonus and enough R&D industries you can hope to have it by second half of 1943. This plane, along with the Jill, will give you the ability to strike from 8 hexes using your CVs...which is basically the only hope of surviving a CV-CV encounter in mid-late war.
Same goes for the Grace. This is, imho, a great plane and it should be R&D heavily.
After having studied the subject for a long time, my overall view is the following:
Use the Rufe to R&D the A6M5 and start R&D the A6M3a (think 3 factories for the A6M5 and 2 for the A6M3a with the engine bonus are more than enough). As soon as you get these models, abbandon the whole Zero line and convert these 5 factories to the SAM line.
I still think investing any resource in the Shidens or in the Jet-planes is a waste of resources