We have come to believe that the problem is that the bond to the paint is just so strong that there is no reasonable way of breaking it. Also, the coating can be harder than the paint and that 'scratch resistance' which we like actually means that a polish/compound has less effect on the coating than the paint. So, as you polish it, it is a bit of a lottery as to where you first breach the coating. once you do, you are going to start eating through the paint quickly whilst the coating in the surrounding area is still resisting. Our experience has been that this can mean you start getting a rather rippled, orange peel type effect.
We have yet to explore this in great depth and this may be a result of having used heavy compounds and it may not be the same with something light (though, you would need to do a heck of a lot of work, in this case).
This is all something of an educated guess as it is something we have only just discovered. It is also plausible that the solvents (which are pretty potent) have modified the clearcoat such that, once you remove the coating, what remains is different from what it began. Ultimately I just wonder how much testing has been done with coatings and attempts to remove fully - given that some are intended as permanent, it would not be surprising to find that limited amounts of work will have been done to try to remove said coating.
