I have the same view as a scientist and engineer. I am sceptical about some of the Prof's claims and degree of scientific research. His focus on calories shows a lack of understanding of how the body's metabolism works. The calorie value of food has little relationship to it's energy potential in the body - for example carbs and fats go thru a completely different process and some of fat calorie value is 'lost'. So this '800' calorie thing is very unscientific.
I've clicked 'agree' with that, as in I agree that fixating on 800 calories and ignoring where those come from is not great science.
I see the 800 calorie thing in the same light as the 15kg thing, and suspect they are numbers that come up regularly just to push the message.
There's no doubt that Roy Taylor feels that fat loss at any rate is a potential road to remission, and I've heard it direct from his horses mouth(!) that a specific calorie intake is not important, so I don't think he's guilty of delusion, just over-simplification, especially when it comes to interacting with the media.
And to be fair, in broad terms, 800 cals restriction is going to result in weight loss for most people in the short term regardless of where the calories come from.
I really wish, however, the DiRECT study had investigated various diet styles. Just off the top of my head, if I was conducting the study, I might do this for example:
1 group doing as they did - 800 cals with HCLF shakes plus veg for a period then stepped re-introduction to standard food with the instruction to try to avoid weight regain.
1 group doing the weight loss part the same way - 800 cals with HCLF shakes plus veg for a period, but then stepped introduction of LCHF food with the instruction to try to avoid weight regain.
1 group doing the weight loss part the same way - 800 cals with HCLF shakes plus veg for a period, but then stepped introduction of LCHF food with the instruction to eat however much of it they want, ignoring calories and weight gain.
1 group who didn't previously eat LCHF, switching to LCHF and eating however much of it they want. Weight loss and calories not important, right from the start.
Monitor the weights, HbA1cs and glucose tolerance levels of participants every 3 months.
The only reason I think that the weight loss part can be the same for the first 3 groups is just because it's so radical that it's likely to cause weight loss for everyone, and the longer-term diet is the more important one anyway.
I don't think there would be too much noise in that data to spot meaningful differences. There would be the same problem as the DiRECT study had - you don't know how well people will be sticking to their plans.
It might also be nice to sub-divide the first 2 groups (i.e. the ones where long-term weight regain is actively avoided) one more level: half left to get on with avoiding weight regain all by themselves, and half given regular access to sessions with practice nurses for moral support, guidance on diet and feedback as to how they are doing.
This access to long-term support was something the DiRECT study thought was important, and it would be nice to have a measure of how effective it is.