This is more or less my take as well. Beside the issues that you highlighted, the thing that really doesn't pass the sniff test for me is the completely unsupp

This is more or less my take as well. Beside the issues that you highlighted, the thing that really doesn't pass the sniff test for me is the completely unsupported claim that with SORNs "Uncertainty grows linearly in general". I simply don't see how this can possibly be the case; interval growth in arithmetic is a straightforward consequence of finite precision, independent of what representation is used. The only way you can avoid it is for representations to grow arbitrarily, which introduces its own set of problems.