Mitsubishi built many types of aircraft during World War II, aside from the Zero.
Ki-46 Dinah - Not Likely
The MN 三菱５１９４号 probably is not the number of a Ki-46 because of the characteristics of the stencil. The Japanese army usually used only the number (5194) on stenciled presentations of MNs. So I rule out the Ki-46. There was a Ki-46-III 5194. It was assembled in August 1944. But, as I say, I think you can eliminate it as possibly being the plane of your inquiry.
Ki-51 Sonia - Impossible
The Ki-51 SONIA is out because no individuals, other than the first 11 experimental planes, had four-digit MNs starting with a 5. Also the comments about the presentations of MNs on army aircraft applies.
Ki-21 Sally - Impossible
The Ki-21 SALLY is eliminated by the fact that (1) no models of it had four-digit MNs that started with a 5 and (2) by the characteristics of MN presentations on army aircraft that are at odds with the markings on the relic, as mentioned above.
G3M Nell - Pre-Pacific War Possible
I believe that you can rule out the G3M, as well. A G3M with the MN 三菱５１９４号 was a pre-Pacific War attack plane that was built during July-September 1938. But I must say that it is remotely possible that NELL No. 5194 could have been around, perhaps being used as a transport. But I wouldn't put much stock in that possibility.
G4M1 Betty - Likely
The G4M is a possibility. A G4M1 with that MN was assembled during January 1942. As a matter of fact, Allied forces captured G4M1 Betty 5194 captured at Lae Airfield, New Guinea, and crash inspectors reported the capture in Crashed Enemy Aircraft Report No. 17 (CEAR17). They estimated that the plane's assembly date was in January 1942.
Candidate
Your candidate is a BETTY. And the very number given on the relic has shown up in a WWII intelligence document. I must caution you not to jump to conclusions. Just because it might be a Betty doesn't make it a fact.
