My claim that religious statements are not truth valuable in the sense that logical or empirical propositions are is, in fact a claim that logical positivists like Freddie Ayer, Ernst Nagel, Rudolph Carnap and other have long held. But now a number of researchers (myself, Ara Norenzayan, Pascal Boyer, Dan Sperber, Justin Barrett, Harvey Whitehouse, Robert McCauley, E. Thomas Lawson, and dozens of authors) have shown this to be experimentally the case again, and again, and again (in journals such as Behavioral and Brain Sciences, Nature, Cognition, Cognitive Science, Journal of Cognition and Culture, Evolutionary Psychology, etc.). It matters little what someone’s intuition about this is. If science had to rely on intuition we’d be back, if not in the caves, at least in the time of witches and demons. As Hobbes noted (and Galileo and Aristotle before him) how can one possibly even make sense of “He is one in three” or of a non-sentient being that sees, hears, and smells all, or of even life after death? The way you do it is give it a particular propositional interpretation in a particular context (which is why there are weekly sermons, and also why religious quasi-propositions - which have the syntactic form of a proposition but which are semantically senseless, like colorless green ideas that sleep furiously - survive over centuries when contexts change).
As for people killing out of religious motivation, it happens all the time, just as doing good out of religious motivation happens all the time. Sometimes people do very bad or very good things in a religious cause. But I know of no empirical study or set of evidence that has a scientific control (say of non-religious ideologies) that shows religion is a greater cause of violence than non-religious ideologies.