Your so-called expert seems to be neglecting the fact that the vehicle designers included seat belt stretch when they designed the vehicle. Seat belts are supposed to stretch, as it does increase ride down time. By bracing against a rigid structure such as the dash, you are all but eliminating that component of the vehicle's designed safety system.
Seat belts are supposed to stretch. They could make them with no stretch, or much less stretch, but they don't for a reason. If you look closely at a seat belt, you will see that it is made up of diagonally woven fibers. It is this diagonal weave that gives the seat belt its stretch. If the fibers ran the length of the belt, it would have much less stretch and the passenger would be more tightly coupled to the vehicle chassis. Coupling to the chassis is beneficial to a degree, but there comes a point where there needs to be some give. Seat belt design has accommodated this, by balancing tightly restraining the occupant and also stretching to allow more ride down time.
LATCH/UAS/ISOFIX belts and the internal harnesses on child restraints are also designed the same way. Some stretch is good.
Your expert mentions that many/most forward facing seats exceed the 550 mm standard limit for forward motion. If this is true, how are these seats legal? I know that in the US and Canada, any seat that is found to not comply with the standard would be recalled.
It is also worth noting that the child is NOT "cruis[ing], motion unchecked, forward in space" unless the seat and/or harness were too loose to begin with. If the seat was properly installed with a tight belt, and the harness was properly tightened, the motion is being checked by the seat belt and harness, which are stretching as designed to give the child more time to slow down.
Your expert also touts the tighter attachment of rigid ISOFIX. Again, he is neglecting that many seat manufacturers have designed a crumple zone of sorts into their rigid ISOFIX, because some give is desirable. The stretch of the webbing that was lost is replaced by a crumple zone. If the best performance was to be had by coupling the dummy to the chassis as tightly as possible, why would they intentionally partially de-couple them?
In the US, historically our seats have been made out of flexible plastics. Now we are seeing a trend towards metal frames and metal reinforcement, and at the same time we are hearing from the manufacturers that they are having to add other design components to the seat because the seat is too rigid. Too rigid, no flex, transfers too much energy to the child. And that is exactly what I was talking about earlier, when I said that watching a video doesn't tell the whole story - what were the injury measures experienced by these dummies? I'm sure that in the prototype stage, when these metal-reinforced seats were too stiff and were passing too much energy to the child, the crash test looked just fine but yet the dummies had extreme injuries.