Agree
I agree but more mildly. :=)
Many of the questions seem to be more aimed at how well someone can present himself than whether he'd be useful in the job. Q2 is very common, but the results are almost always indicative, not conclusive. Engineers and software folks are commonly rather poor at expressing this sort of thing, particularly when it comes to blowing their own trumpets.
The skill of the interviewers, follow-on questions, and interpreting the answers is key.
I suspect that the advice given will mainly get you a slick talker, and then you trust to luck whether he can *really* do the job or not.
I remember after an interview thinking that the previous candidate was perfect, and the rest of the panel agreed. When we sat down and really looked at what he'd said and his proven experience, he really didn't have the right stuff at all. We learnt a lot from that one.
Later we found that some of the best candidates were those who argued with us about the best way to do something.
