Is that really the case? I'm not an expert in AG, but I would imagine that any major result published in Weil's framework can either be trivially converted to schemes or it has been ported over by now