I'm not familiar with the Heideggerian critique of cognitivism, or of Hubert Dreyfus' work, but some of your quotes sound agreeable. I am not convinced however that the frame problem and related issues are unsolvable. The way forward is to program, measure and improve.