Do we get to use math in Model B, or just words? How are we allowed to "consider Model A and Model B"? Model A certainly has math associated with it. Does Model B have any math at all--for instance, does Model B accept that mass and rest mass are different things?

For instance, suppose Model C says, "2^(10^1000) is even", but Model D says, "2^(10^1000) is not even". And I might say, "And you can't use any math. You just have to consider C vs D." What, then, are we considering? Without math, we can't really judge anything. We're just left in self-enforced confusion about which statement might make more sense.

(Also, I note that the argument you use against indirect evidence also poses problems for quarks, and neutrons for that matter.)

(I also note that physics is filled with oodles of experiments that don't pick one hypothesis over the other beforehand. The characterization is wrong. However, let's ignore that and consider these models on their own merits.)