One thing I have been arguing for a long time is that the fundamental constants are different until we observe them. i.e. if we don't observe it, it's possible for a tennis ball to travel through a wall. But in the universal program, if we will now or later observe the result, then it won't happen. But it'll happen so long as we will never observe the result. In fact, it's probably happened many times.
Impossible means it does not happen, not that it does not happen only when we look. Just because we can't see it doesn't mean that it doesn't happen. After all, as the comment I replied to pointed out, other galaxies can have different constants. We have to be humble and admit we just don't know.
This seems like a distinction without a difference, since we can never positively categorize any unobserved phenomenon as impossible (vs merely unobservable). To me, it seems ontologically cleaner to treat existence and observability as the same thing. shrug
Okay, fine, I'll come clean, I was just making an unfalsifiability joke. The original god-of-the-gapsy comment was the one that got me. Always just out of reach of our verifiability is the magic. Why not all the way out?
No one has proven that this is impossible, AFAIK.