It's not that the first-order advice of "avoid blue light in the evening" is necessarily bad. Rather, telling people (in no uncertain terms) that if they do <x> they are hurting themselves and that they must change what they do, may cause second-order problems. This is one of the reasons why you should not get medical advice from the internet or pop science books.
(FWIW, I have read the book, before learning of the scientific critiques. I must agree that the presentation of the book is usually much stronger than what the data presented admits. It's not that the small-scale studies in question seem particularly flawed or that the interpretations are downright unreasonable, but I think it's borderline unscientific to present as fact a story woven from more or less isolated data points in the form of individual studies.)
Also, I’d like to actually understand the inaccuracy... it seems the real problem is he “emphasised” some things. That’s hardly inaccurate.