I think that this is most likely an effect of the current Netcode in the Beta. Because the Netcode is so much better, it is much more difficult to distinguish between frame lapses due to online/latency/built in delay issues, versus genuine changes to the frame data that would remain present in an offline version of the game.
Frames in general actually matter a great deal more than they did prior to the GGPO upgrade while playing online, but the Beta in its current form operates with an automatic delay of an unknown number of frames (we know that the final version, when it reaches the actual game, will be set to 3 frames of delay), and that goes along with whatever other issues might take place as a result of Rollback that might cause small skips or discontinuities of some kind that could further muddy the waters when it comes to reliable frame data testing.
Basically, while the new netcode may be terrific in that we can have fun playing the game online, and also in that winners and losers are no longer determined by lag or the scrub tactics devised to exploit it, it is still an incredibly unreliable place to test for any changes in frame data that are not glaringly large, or for accurate frame data in general. While the minor discrepancies between the ways the moves are expressed in terms of frames in the GGPO Netcode versus those of offline are unlikely to alter the outcome of a match, they are more than enough to make the recording of accurate frame data impossible.
Additionally, Jax has received a number of alterations to his frames in the EO Beta, and all of them have been reflected in the movelist thus far. It seems unlikely to me that they would remember to record that D1D2 is now -5 on block instead of -7, and then somehow forget to record a major change to arguably one of his top 3-5 most important moves. I don't think we have anything to worry about, it's just the connection playing tricks on us.
