My take is exactly the opposite: since 3D structures of proteins alone are almost never the bottleneck in drug discovery, this won't actually change anything. Knowing how the drug is going to bind, and knowing how it'll behave in vivo, are not something you can predict from deep sequencing data.