The difficult problems remain difficult, no matter what the context. I perhaps foolishly chose "Data Binding Interop" for my first session to see if anything had changed. I was rewarded with echoes of 2000, with more emphasis on structure and less emphasis on type obsession. Instead of arguments between vendors it was arguments among framework creators. I looked a few times to Jeni Tennison and Norm Walsh for confirmation that we'd been here before.
(I may also have even less sympathy for "data binding" than I used to because too much time in Erlang makes it hard for me ever to see object models as saviors.)