Ive never thought of it in terms of g-onset rate; but if you figure it in terms of alpha-dot or pitch rate then there may be concerns over, perhaps, inertia coupling effects (for the latter) and amybe some kind of downwash lage effect for the former? Of course, it may just be a consequence of gross overshoots of the target g.
Still sounds odd to try to limit the rate at which g can be built, regardless of the actual g level. A rapid change from 1g to 2g on an 8g rated a/c shouldn't be a hazard surely, whether it happens in more or less than 0.1 sec?