If one wanted to draw a distiction--and I agree that the two might be close enough to not bother--it would be that in reciprocal altruism the benefited party reciprocates without regard for whether it is in their self-interest at this point. The overall scheme is in one's self interest: A helps B, B helps A. But at any particular moment, if you're doing the "what if I betray you--would it be better for me?" calculation for someone who has already helped you, it's self-interested sometimes-mutually-beneficial egotism, not reciprocal altruism.

So the armor farming example distinguishes between these two, at least from my understanding of how one would want to use these terms if one wanted to differentiate. In animal behavior, one generally doesn't distinguish, but the presumption is that the sophistication of models needed to be a rational egotist in mutually beneficial situations is considerably higher than is needed for reciprocal altruism, so you generally assume the latter. Or, you don't even bother drawing a distinction--if you're doing it because you think it all through carefully instead of because of a feeling of it being nice to help people who are helpful, that's just a detail of how to implement reciprocal altruism.

(I think it's worth making such a distinction because it has important consequences for understanding the effort we want to put into avoiding betrayal and/or setting things up so betrayal is not even mildly advantageous.)