Read the paper (https://dl.dropboxusercontent.com/u/85192141/2012-scannell.p...), it addresses these objections. The "easy problems" objection is just a form of the "no true Scotsman" fallacy, where you redefine difficult problems to be the ones you haven't stumbled across a solution for.
Uh, isn't that what makes a difficult problem... difficult? I don't think there's any redefinition going on. I'm not sure I've ever heard somebody say, "this is really easy, I can't figure it out."
The thing is we don't figure out drugs, we find them by inspired accident. The attempt to move to a "let's figure it out" approach with computational methods seems to be what's making us worse at drug discovery.
