Experiments in any science are often spotty at first because the phenomenon are not well understood and the experiments are not well designed.
That's probably the key part. Similar to what was said before, I'd ask for specifics on experiments being "spotty" ...it doesn't seem like this has really been the case in the solid sciences if you think about it. I almost get the impression psi-experiments are being referred to as if they've been confirmed now. In that sense it's almost begging the question.
Experiments aren't very spotty when they focus on falsification and deliver it. But in considering the objects falling at the same rate idea, I get what he means in that aerodynamics had to be factored out. But still...the thing is, if you have a problem with the experiment state it. Otherwise go home. What does it sound like to go "I know I'm right, therefore the experiment must be flawed"? Right, sour grapes.
This is post-hoc rationalization.
You shouldn't be starting out believing something then trying to prove it, you should be just having an idea and trying to devise a way to determine if it's true or not -or as science really works, devising a way to prove it false, and what result in the experiment would show it's false.
In the case of prayer, when you really think about it, there aren't many factors to fiddle around with...though "god knows" people have tried. "Make them pray this way...no that way" When come on....all knowing God? How then can you not get the message through, (and shouldn't the mere act of it being a study on if it works give god all the more incentive to heal the shit out of those people?).
Just remember, if someone has real grievances with a study, say what they are...otherwise you could use this line about anything out of hand without even looking at the details. The simple truth is, they just have a belief in their minds beforehand and are unwilling to give it up in light of evidence to the contrary Instead they do these sort of post-hoc rationalizations.
It's basically "So instead of finding fault with the experiment, you're going to dismiss it anyway because something must be missing because you know you're right?" After that I might say "How then could you know you're wrong?"
Anyway sorry if I'm repeating myself, it's gotten very late.