Basically, he's saying existing GW-scale reactors are too big NOT to fail (at least in various loss-of-cooling scenarios). I guess whether that makes them "flawed" depends on one's faith in always being able to come up with a way to maintain cooling in an accident, or never to have such accidents in the first place. Obviously at Fukushima such faith proved unwarranted!
His overall recommendation sounds like it could be music to the ears of the small modular reactor crowd. And it's also clear that he's advocating a phaseout and not wholesale immediate shutdowns. I'd certainly prefer to see reactors with newer technology rather than extending licenses out as far as 60 and 80 years.
And of course, that leaves aside the question of whether we can do without nuclear; my last remark is just my gut feeling regarding which of two alternatives for fission I'd prefer. What role fission should play in the overall energy mix is less clear to me.