Good point.
I guess my problem is that it takes so long for the experimental support to catch up that sometimes it seems like it will never happen. Also, some of the stuff is so wacky, that an experiment does even look feasable. So, its easy to just say, that can't be possible, rather than sit back and wait for the 100's of years it might take for the experiment to be performed to verify the math.
So, today we are verifying some stuff that was really wacky back in the day. How much of that wacky stuff do you think was false? If only a fraction of it has been shown to represent reality, with the rest being crap, then maybe it isn't cost effective to get too wacky.
Sorry to be so vague, I hope you get my drift.