I appreciate that he has been involved in hundreds of usability tests and watched thousands of people use websites. (Although I don't think he has actually designed any). But he still needs to back up what he says with the relevant data. Too often his Alertbox email is full of bold proclamations which sound like those from a fundamentalist, rather than a considered and balanced opinion.
I think what turned me off to him was his proclamation a few years ago that flash was 99% bad. He has since modified his views slightly, but they remain pretty vitriolic.
What he has started to get at - albeit several years later - is that 99% of Flash designers are bad.
Certainly recent innovations in web development like increased recognition of the DOM and the use of AJAX (wonderful and welcome as they are), achieve effects that have long been achievable using Flash.
An example of something outdated - well, perhaps 'outdated' is the wrong word - but his preference for liquid layouts over fixed layouts, is baffling. I agree that liquid layouts are a definite boon (my own site uses a liquid layout) in some environments, but to state 'On big monitors, websites are difficult to use if they don't resize with the window', is just not the case. Badly designed websites that don't resize are difficult to use, I agree.
An example of something that's just plain wrong: as recently as 2004 he stated that cross browser compatibility for websites was a 'one star guideline - worth thinking about if you have extra project time, but not a priority'. Cross browser compatibility is something that should be near the top of every designers concerns - especially those concerned with accessibility. Designing for text readers - rather than visual browsers - is of key importance to, say, blind people using the web.
Again, I agree with an awful lot of what he says and without question, every web designer can learn *an awful lot* from him - my principal problem with him is that everything he says is taken as 'The Law' and that if he has said it 'then it must be true'. I think that his status can discourage people from investigating design choices that may well be appropriate.
