Pushing model builders to use smarter scrapers is a net good. Endless rescrapes of static content is driving up bandwidth bills for housing simple things.

This will lead to (if anything at all) smarter input parsers, not smarter scrapers.