Since you showed up on a Google alert (I co-founded NuSI with Peter) I'll respond.
NuSI funded two trials that could test the energy balance hypothesis. (This is an important point: these experiments tested the null hypothesis, which was energy balance, although the experimental tests were based on predictions made by the carb insulin model.) One trial, led by Kevin Hall and Eric Ravussin, was a non-randomized pilot study that they interpreted as consistent with the energy balance hypothesis. That interpretation was challenged by David Ludwig who led the other trial, an RCT that Ludwig interpreted as inconsistent with the energy balance, calorie-is-a-calorie model. Hall et al challenged that interpretation.
If Peter changed his mind, which his book suggests he did, it could not have been based on the results from the NuSI-funded trials because their interpretations were contradictory, as experiments in science often are. Peter cares deeply about scientific method so I'm sure he understands this.
NuSI funded two trials that could test the energy balance hypothesis. (This is an important point: these experiments tested the null hypothesis, which was energy balance, although the experimental tests were based on predictions made by the carb insulin model.) One trial, led by Kevin Hall and Eric Ravussin, was a non-randomized pilot study that they interpreted as consistent with the energy balance hypothesis. That interpretation was challenged by David Ludwig who led the other trial, an RCT that Ludwig interpreted as inconsistent with the energy balance, calorie-is-a-calorie model. Hall et al challenged that interpretation.
If Peter changed his mind, which his book suggests he did, it could not have been based on the results from the NuSI-funded trials because their interpretations were contradictory, as experiments in science often are. Peter cares deeply about scientific method so I'm sure he understands this.