[chrisseaton]

I think most of the other papers have non-experimental proofs, so it's not that they do nothing rigorous.

And anyway the claim was 'very rarely', when it's actually a majority of cases.