This is just a guess, but:
This Usenet posting presents what it claims to be the XTOL formula, based on Kodak patent information. Although I wouldn't bet money that it really is the production XTOL, it's probably in the right general ballpark. The Usenet formula has sodium sulfite in both the A and the B packets, and in fact that's the largest single component of (the Usenet version of) XTOL by mass (prior to mixing in water, anyhow), with a total of 10g in part A and 75g in part B per liter (50g and 375g, adjusted for the 5-liter packets you showed). Thus, my guess is that Kodak decided to shift some of the sodium sulfite from packet B into packet A or vice-versa, depending on when the two XTOL products were bought. I've no idea why they would do this, although I could speculate (improving reliability, improving mixing characteristics, reducing manufacturing costs, etc.). If I'm right, then the final mixed product would be identical in either case. (I note a 1.7g discrepancy in the total masses on the packages you've got, though. If this isn't just rounding error, it could signal some additional subtle change in the XTOL formula.) Such changes in manufactured products are hardly novel. There's been a lot of talk here on APUG recently about Rodinal and its history, for instance, which has included (poorly understood) changes in Rodinal composition over the years.
If you're concerned, you might consider calling Kodak and asking about it.