At the the risk of speaking for the OP, I suspect what he is getting at is that in real-world practical scenarios, there are things that masses can do to gravity for which simple acceleration does not (normally) have an equivalent. Such as, say, "wobble".
And that this could be used to
practically tell the difference between gravity and acceleration of a
real-world event being measured.
roineust , I think you're missing the point. It's a
principle - that term implies that you can (in principle) eliminate
practical factors that might
pollute it.
Detecting a "wobble" in the force you're measuring and being able to declare "it's gravity, not acceleration" doesn't invalidate the
principle - any more than detecting tidal forces does.
Tidal forces and wobbles are examples of practical factors that
pollute the ability to accurately
test the principle; they don't invalidate it.
A spurious example:
Objects near Earth fall at 9.8m/s^2. That's hard to test that
if you have no way to get rid of air resistance in your tests. That doesn't
invalidate the known free fall acceleration near Earth - it doesn't mean it's falsified.