In particular, I'm saying Pr(calcsam experiences warm feelings after reading the Book of Mormon | LDS church is true, social interaction with members) > Pr(warm, glowy feelings | LDS church is false, social interaction with members).

I agree that it's really difficult to say exactly what these probabilities are. If you forced me to assign numbers, I would assign something close to 1 for the former and .1-.8 for the latter. To be valid, these should really be the result of probability flows through an entire network of beliefs, but I think the direction of the inequality is apparent. I agree that spiritual experiences for different religions will tend to offset one another. Whether these in total constitute net positive or negative evidence for a god in general depends of your prior about how these experiences are distributed. Based on my interpretation of LDS doctrine, many non-LDS individuals would feel the spirit, even during non-LDS services, just not as strongly.

In any other forum but this one, I wouldn't go around saying this is evidence, but it is, if only weakly.