The biggest unaddressed problem I see with Ethereum is the potential of the blockchain growing very large. There has been some discussion of this (on the Ethereum blog and elsewhere), but these discussions focus on showing that the blockchain will remain secure at very large sizes.
I'm more interested in the operational mechanics of having to deal with a very large block chain. The figure of 100 TB was used in one blog post, and that is a challenging amount of data to deal with for anyone.
What I haven't seen is any progress on any proposals to either provide incentives to limit the size, or ideas around splitting the chain.