I had this problem a lot, particularly when coating glass coverslips. For me, the problem was completely solved, when I switched from laminin "purified" from tutors to recombinant human laminin. That is a bit more expensive, but I could skip the PDL coating step and it worked out.
We culture NPCs on both mouse and human Laminins (10-20 ug/ml) and on Geltrex (thin gel coating method) with good success. Both need to be freshly made (no freeze-thaw of stocks) and coated plates are best used with a week or so. I personally do not precoat plasticware with POL, and for seeding on glass use extended rinses before LAM coating.
NPCs need to be handled as gentle and as fast as possible and the coating material should never get even touch-dry. Sometimes addition of ROCK inhibitor can help with initial attachment after seeding.