Last weekend I presented at a local conference, with faculty members from various departments across campus talking about their work. Since I have been working on a paper (with Maarten Boudry) having to do with the limits of physics and how we might find a signature of a supernatural cause in data, I talked about that—nice interdisciplinary topic and all that.

1. Some see this as an unacceptable putting God to the test. (Though I talk about supernatural agents in general, not any specific deity.) We end up with something akin to a conspiracy theory: If you test God (or approach him with less than fully pious attitude), he will make sure your test is useless. 
2. There’s also a more liberal-ish religious response. Since God is supposed to be metaphysical rather than physical, or because God is “the ground of Being” or for some such reason not an object among other objects, evidence has no bearing on the matter.
3. Some atheists object, always preferring a naturalistic explanation of any evidence over any claim of supernatural agency. And there always is a naturalistic alternative. (Just like there is always a supernatural alternative: Satan planted all those fossils!) 
4. Many, regardless of personal position, find the notion of testing supernatural claims distasteful, since it breaches a firewall they imagine exists between science and religion. Playing around with empirical data—science—can never say anything about the supernatural. Possibly because a controlled experiment cannot be ensured or some such reason.