The answer to [this question](https://physics.stackexchange.com/q/61243/) and the comments on it implies that general relativity has not been rigorously shown to be non-renormalizable for all loop diagrams -- only shown for two loops. However, doesn't [this paper](http://arxiv.org/abs/0709.3555) give a complete proof that GR is non-renormalizable without referring to the two-loop computation mentioned in the question's answer?