It's important to remember that Pathfinder was starting from a very different place than D&DN. Pathfinder was trying to evolve 3.5 in the same general way that 3.5 evolved from 3.0. It wasn't trying to create an entirely new version of D&D. D&DN is a much more involved process where they are really trying to incorporate the strengths of various editions into a single, more flexible version of the game. D&DN needs alpha testing in a way that Pathfinder simply doesn't.
That being said, I agree with the general principle that D&DN needs vastly more testing than its seen so far, and more testing than I think it's likely to see before release. As much as WotC took a lot of flack for publishing 3.5 so shortly after 3.0, it was a much better version of the game. Likewise, I think 4e would have been much stronger if they had started out with the lessons they had learned by the time they released Essentials. After 3 months of playing 4e, my biggest wish was that WotC had tested it for another year. I appreciate 4e errata, but it's annoying that my books are so useless.
I sympathize with the problem that WotC can't just release an "almost done" free version of the game and then expect everyone to purchase the final version. But at the same time, I wish they had a better mechanism to get wider feedback and more polish before they release the "final" version of D&DN. Speaking just for myself, I'd would gladly pay for an "escalation" edition (to borrow from 13th Age) that came with a series of PDFs of increasing "doneness" and ended with receiving a physical copy of the completed rules -- especially if those rules were better tested than 3.0 or 4e.
It just takes a few years of solid play to knock out all the bugs in a game like D&D. I really wish WotC had a business model compatible with that reality.
-KS