It's simply the difference between claiming something is "impossible based on our current understanding/model of physics" and "impossible regardless of what we

It's simply the difference between claiming something is "impossible based on our current understanding/model of physics" and "impossible regardless of what we don't understand". Science gives us the first one, but the original commentator unambiguously claimed the second. Our models of the universe being wrong happens all the time, sometimes in significant ways, and even now there are plenty of unexplained things in our universe that have the potential to change our understanding. It seems pretty optimistic to assume our current models are guaranteed correct in every circumstance we haven't yet observed even when they can't explain everything we have actually observed. So yes, perpetual motion could start working tomorrow, or could already exist for all we know. That doesn't mean it's wrong to say physics _currently_ says it's impossible and any claim better have some pretty good evidence to dispute it (which none currently do). But that also doesn't mean that if we're around centuries/millennia from now we won't have found a special case where our current understanding breaks down.