In a rough sense I agree with this. I discussed earlier (with Kwayera I believe) the issues involved in adhering to a straight out verificationism (/logical positivism) which asserts:
(1) "The only meaningful claims are those which are verifiable or falsifiable."
A possible corollary of this would be:
(2) "It is only reasonable to believe in those claims which are verifiable or whose negation is falsifiable."
But then a problem arises. Neither (1) nor (2) are verifiable/falsifiable, so we need to have at least
some beliefs which fall outside the purview of hardline empiricism. How are we then to judge the reasonableness of beliefs falling into this extended realm? This is by no means an easy question, and it is quite possible that no clearly correct answer will be forthcoming (it is partly for this reason that some people are driven to pragmatism - that we should accept those beliefs which prove useful). This issue also borders on the topic of 'properly basic beliefs' that we have dealt with previously because some people think that properly basic beliefs constitute the best starting point. However, even if you can't completely reject unfalsifiable claims on empirical grounds you need a pretty damn good reason to admit it to the
a priori knowledge club. If nothing else, Ockham's razor (a long term member, it would seem) does short work with those entities whose existence is unsupported.