Phinney is awesome. He’s probably the most officially legit guy we have in some respects. I see nothing in his video that seems inaccurate.
Something worth understanding. I have an interest in a different science field and got to know the leading researcher, a physicist. I found it interesting because despite he sees the individual’s reality of the topic, he is adamantly about “the data.” He will not say anything that does not have data to support it. He will not scoff at new ideas but he will insist that rather than running away with enthusiasm we recognize that there is not data to support it. I’ve never met anybody so absolutely “grounded in research” as this guy.
Phinney is like that. Nothing he said in that talk contradicted fasting. He didn’t say there was ‘no benefit’ he said here is the data about known issues with it. And he pointed out up front not only that it did have some known benefits (short term) but – to me this is the MOST IMPORTANT question – that we do not yet have PR research documenting what happens AFTER long term fasts – how much of that metabolic reduction, that nitrogen (protein) loss, is recovered? – and that could make all the difference. Maybe the body recovers that, and maybe it recovers even more. That would change the whole picture.
(Edited to add: there is always LBM lost in fasting, even after recovery. But there is a possibility that, like a corporate layoff, it’s for the best. As long as it’s not too extreme, and fat loss is larger, and the metabolism is not damaged after-recovery. The Biggest Loser, of course, was calorie restriction, and it’s the only ‘data’ he had on this, but it’s not close enough for me to consider avoiding fasting.)
Refeeding syndrome is major and the longer the fast and the worse it’s done the more it can hurt, seriously injure, and even kill people. It deserves more attention and much more emphasis with guidelines. It’s by far the most dangerous part of extended fasting.
It’s said on Fung’s forum that he recommends 10 days “because that gets the best compliance” (the forum officially does not support any longer fast though I see mentions of others) although he sometimes assigns 14 for the seriously ill. I suspect that aside from compliance it also reduces the chance that refeeding will kill someone and indirectly wreck his life.
I read a book written in 1910-1911 by Upton Sinclair, and a book written in 1934 (rev 1950) by Herbert Shelton. Dr. Shelton had supervised over 25,000 fasts at the point where he wrote that! And several other docs he refers to several times in the book also had large experience with this. They did it for health, only minimally for fat loss, since the number of people fat (and the degree) in 1934 was pretty dramatically less than now. His fasts were long and expected physical Herx symptoms in major ways. Although the book got a little boring in the middle (repetitive) because he was rebutting many popular claims, critics, and others he felt were inaccurate of the time – which we don’t care about now – still on the whole it’s a great book (cheap on kindle) and presents a much more serious, health-based, and detailed review of fasting than I think most of the modern enthusiasm includes. In part because nowdays any doc has to worry about getting sued so aside from ‘blood sugar’ ‘obesity’ ‘autophagy’ they can’t say much, like the gradual autophagy-effects on body systems and parts that are unwell (but which takes real time, sometimes 1-2 (in rare cases 3) months). It also focused on a ‘complete’ fast which means until the body finishes its cycle, the tongue clears and the ‘genuine hunger’ returns (we are talking, generally, six days to 1-2 months).
Try to talk to anybody anywhere about extended fasts and mostly all you get is moralizing about why you shouldn’t. Even articles online devolve into that. My point is only that the world of fasting in 1910 and 1934 and 1950 was pretty extensive and had a rather different approach and outlook than what we have now. It’s worth seeing the difference, if you’re really interested in fasting.
Phinney’s good. He stays with the data. We need researchers like that. We also need people with enthusiasm for pushing the boundaries and driving the layman experiments and then the research experiments on new stuff. There’s room for everyone. This field is niche of niche as it is. Not exactly overcrowded. 
(Ugh.)