For organizational purposes, if they were to do a Pathfinder 2.0, they could make PFS rules (which judging from posts apply to a fairly large fraction of their market) be the Core Rulebook 2.0 rules, but not including the campaign setting-specific stuff, and then release supplements that broaden the horizons. This would have the advantage of making the Core Rulebook 2.0 easier to get out the door (since whatever they DON'T change going from 1.0 to 2.0 is well tested in an organized situation), and then other stuff could be tweaked separately (taking a longer time. This would also have the advantage of making the new Core stuff a lot easier to find, even for non-PFS people (who would also need a one or more supplements, but not so many as before, at least until a similar amount of time passes compared to the time between the original Core Rulebook and now, for future rules bloat to spread everything out again). If the transition from 1.0 to 2.0 was like AD&D 1st Edition to 2nd Edition (or at least 3.0 to 3.5), people could even keep using the stuff they already have, although I would expect that some people would buy the new Core even when they already have most of what it covers just for organizational purposes (and then the supplements that come out later would offer both reorganization and bug fixes, as well as new material, and thus have an attraction of their own).
The CRB is a hot mess currently, lacking definitions of key game terms (there is no definition for Burrow, for example) and including redundant and conflicting definitions of others (movement-impacting things, like difficult terrain). IMO, they could save multiple pages of wordcount by having a chapter glossary to start each chapter. Numenera was fabulous for including not only key game terms, but also margin notes with references for key game concepts throughout the entire book.
Except it only really works in the PFS context: strictly defined goals per session.
Very hard to apply in a more flexible game. You could just say "Every three sessions", but otherwise you'll have to railroad specific xp goals in.
I co-wrote a game some years ago where there were 40 character levels and the rule was "you level up once after each session unless the GM says otherwise." Those levels were significantly less choice-laden, mind you, but the same principle applies: how effective is your storytelling if you can't finish it in 39 sessions? In PFS-style experience, 39 sessions = level 14. Capping at 20 is an additional 18 sessions, totaling 57. That's weekly for over a year. If you still feel constrained, you tell players that you'll be slow-tracking experience, resulting in up to 114 sessions. For reference, on full slow-track, you'd see level 10 after 54 sessions.
That is a lot of sessions. Even in a less-linear, more flexible game, I would expect that many players would be chomping at the bit for more mechanical growth after 6 sessions per level for 2 levels. I certainly don't want to be one-shot by an orc warrior after 5 sessions of character building! That's what you'd still be looking at in slow-track.
GMs would still have the option of fiat experience by targeting objectives, which many GMs already do: "You found the maguffin, you level up." Again, the presumption here is about what's printed, not what's foisted upon you as poor GMs and players.