Let’s say baseline is 120bpm (for calculation purposes). With a 0.05% deviation of this beat count, after one minute the track will have drifted + or - 0,06 beats. Assuming you are low on one deck and high on the other, the maximum would be 0,12 beats. As one beat takes 1/120th of a minute, which is 2 beats per second or 1 beat per 0,5 second. 0,12 times 0,5 equals 0.06 seconds or 6ms. If one deck is on track, then it would obviously be 3ms. Again, after a full minute.
You’ll say that 3ms is three times as much as 1 ms. True.
But the next best step is not 0.00 but 0.02. Difference to 0.05 % = 0.03 %.
At 0.02% the difference would be + or - 0,048 (roughly 0,05). 0,048 times 0.5 = 2.88 ms. One deck on track that makes 1.44. Rounded up to 1.5ms. My quick and dirty of 1 ms wasn’t too far of the mark. While relatively still 1.5 times as much (almost), in absolute sense losing 0,48 millisecond a minute is still, as I suggested earlier, nothing,
After all, this is only relevant when manually beat-matching and only with (computer) tight tracks. Anything other than that will have beat drifts that are bigger than 0.05% anyway. And if you are using sync, the whole discussion is mute anyway as both tracks will stay in sync and move up/down in 0.05% increments together.
Granted, if it is technically possible, there is no real reason to not have the option (like the suggestion of having a utility setting). My point was more about the actual “hurt” of this change.
At the end of the day you have to do/use what works for you.