Test Not Worth Conducting

It would be an interesting experiment, but this is what would happen ... if the result was statistically in favour of a prayer-granting God then non-believers would claim that more data points are necessary. If the result was God-neutral then believers would claim that God knew about the test and didn't want to demean Himself by getting involved.

Your concluding point is spot on: there would be no point :)

There are lots of troubling aspect of such a test:

• We wouldn't be able to really discern the true faith of those in the "believer" control group. Can we know that someone who prays with a healthy dose of unbelief underlying their prayers is as effective as those who are true saints?
• It would be difficult to know if what people say they're praying for is what's really going on in their hearts.

Plus, I'm thinking God would look down upon such a test and think, "You poor creatures. You just don't get it, do you? I don't want you to try to prove something like prayer. I'd rather you just pray. Come to me in fellowship ... don't turn it into a science project."

And, as you say, regardless of the results, nobody is going to be happy. :-)