I think the suggestion is, when there is not great evidence either way we should not support or shutdown ideas within the realm of probability - let's be honest and simply say "we don't know".