I suspect that many have been using it for some timeIt would be very interesting to learn who has been using it, for how long, and how they stumbled upon it.
I always get suspicious when someone holds a big lead in a high-scoring golf -- as mtve did at sha-256-hashing, leading 396 to 433. So I asked him about it directly. And he then (finally) remembered ... why yes he did use it in that game! You see, he'd completely forgotten about that when I contacted him several weeks earlier to help me research the Compression in golf series! Well, it was years ago. In fact, mtve knew about this technique since 2002 because he co-refereed TPR(0,6) with Ton Hospel and so was privy to the reason behind Ton's new "At least 55% of the code must consist of normal ASCII characters" rule.
I also wonder about Ruby golfing phenomenon flagitious. After all, he led that same sha-256 golf in Ruby by 338 to 386 for years ... until "dmd" recently wrested the lead at 335 by using "pack u" compression. Now dmd is such a strong golfer that if he was using compression and flagitious was not, I would expect dmd to be at least 20 strokes ahead. Whoa, I just realised that I tied flagitious in Ruby at 173 strokes in the 99 bottles of beer game without using compression ... so if flagitious did use compression in that game, that would have to rate as my greatest Ruby golfing accomplishment. :)