I've downloaded it and started looking it over. Some of the changes I've seen so far, like giving a character an extra hit point per level of his or her favored class and the fighter's armor and weapons training I like. I don't like the fact that humans have to choose a single favored class at first level, and I think the armor and weapons mastery go too far. I don't at all like raising the hit dice of rangers, rogues, sorcerers, and wizards. I don't see much need to change the way the game handles first-level hit points. I'd thought about combined a couple pair of skills into single skills, although I would have combined Balance and Tumble, which both use Dexterity, but not Jump, which uses Strength, into Acrobatics. (Yup, I had the same name in mind.) I'd also considered combining Listen and Spot (both Wisdom-based) into Perception. (Again I had the same name in mind.) I use the Singin' Sheet to do many of my calculations for me, however, so I'd have to give it up and do them all by hand again if I were to make such skill combinations. In any case I wouldn't want to include Jump under Acrobatics, transforming it from a Strength-based ability to a Dexterity-based ability. Jump is one of the few skills at which fighters and barbarians can currently excel, while rogues can already excel at so many skills, so that Jump should be kept separate and Strength-based for the benefit of fighters and barbarians.
For clerics, I see no particular need for at-will orisons versus the current system of a limited number of 0-level spells, especially at the cost of losing cure minor wounds, which PCs and NPCs use all the time for such things as reviving unconscious NPCs for interrogation while keeping them disabled. While the domain powers add a nice touch of flavor, I'm not sure that it's worth reducing the flexibility of spell choice. I strongly oppose having the turn undead ability grant healing to living creatures, both friend and foe. If you want to use the power of undead turning to heal creatures (and it should be at the choice of the cleric, not all living creatures) then make a feat along the lines of those in Complete Divine.
I don't mind having a slow, a moderate and a fast advancement experience progression, but one of them (the slow one presumably) should correspond to the current advancement. Using any of your advancement progressions would slash the level of all the characters in the party I currently run. After the players have done all the work to get so far, slashing their levels is the last thing I'd do. Nor am I willing to extend their current levels through much higher experience requirements to the next levels. I have to add that I'm disappointed to see nothing in the pdf for epic levels. I'm hoping that eventually you'll add epic material, but given the way you've ratcheted up the experience progression for the higher levels, it seems more likely that you intend to deal with epic levels by making it impossible to ever get enough experience to get there.
Based on what I've seen so far--and I haven't yet read everything in the alpha pdf--it's more likely that I'd adopt an item or three out of the rules rather than adopt them wholesale, making it less likely that I'd purchase the books (beta or final) when they come out. The further the rules depart from 3.5, the less likely I'll use any of them. Items like an extra hit point per level in the favored class don't require much adjustment, but I don't really want to sit down for each and every new clerical domain I've created and have to create new domain powers, or for each item like weapon mastery's 100% chance of avoiding being disarmed have to house-rule that it simply grants a +20 or +30 bonus on checks to avoid being disarmed. I'm already doing the latter sort of thing with spells like freedom of movement, true seeing and mind blank, and if Pazio's Pathfinder requires me to do the same or even more, I might as well stick with 3.5.