Advancements in either p2pool software or something similar would go a long way in starting to alleviate the problems associated with mining centralization. Unfortunately, there hasn't been enough incentive to advance these things. I think instead of removing mining from the reference client, it should have adopted and improved upon the p2pool software.
It's not from lack of incentive. Many of us have spent a lot of man hours with discussion, theories and code in mining that the non-mining world probably is not remotely aware of and quick to dismiss as "greedy miners." The idea behind p2pool was sound, but the reality was that the design is fundamentally flawed in a way that cannot be fixed, making it worse to mine on than a regular pool. P2pool is ultimately simply a merged mined blockchain on top of the bitcoin blockchain - mining on the merged chain means you're simply solo mining on a chain with slightly lower difficulty, yet still far too high a diff. After extensive discussion and investigation it is clear that these flaws cannot be fixed to make it even as attractive as regular pooled mining, let alone better. The design of the current bitcoin proof of work itself means that will always be the case. Without a massive change to the blockchain and proof of work design, pooled mining will always be possible, and the "p2pool" design will never be as good. A distributed peer to peer proof of work design that intrinsically does not lend itself to pooled mining without even changing from sha256d allowing existing hardware to continue mining is indeed a solution but unfortunately p2pool is not it and cannot be made to be it.