Listening to some old podcasts, I stumbled over the idea of "failing forward" mentioned in a March 2020 (in this interview) edition of Court Games (RPG). The game designer Max Brooke there described the idea like this (starting at about 1:03:00) when talking about failing unexpectedly with Korvar, one of the hosts:
Max Brooke: Another thing I'll do is [...] let them succeed at [a] cost. We don't have a hard succeed-at-cost mechanic, but sometimes it is narratively more satisfying and fitting. [...] You succeed but you might not like what happens.
Korvar: Yeah. Actually, you have over succeeded. You haven't failed. You've over succeeded. [...] But anyway, you can kind of do that with a dramatically unsatisfying failure where it's like you succeed, but you might not like what happens. Yeah, I think actually a succeed with cost thing, the kind of fail forward. I like that as a mechanism, and it's obviously quite a good thing, too.
Then, I remembered that I believe Exalted 3rd Edition (2016) had something that was described to me as "failing forward" in some manner when I was encountering it the first time. Well, looking at the book again it is more a call to be an absolute fan of the PCs and say "YES" to all things the PCs do, or at least never say outright "NO" but instead answer with some kind of compromise in the shape of "No/Yes, but...".
So, I presume that at some time between 2012 and 2020, or likely even before, Failing Forward had become a specific concept in game design... but when, from where, and what was the idea expressed with it?