It's embarrassing how much some people seem to forget about rationality and skepticism when thinking about this issue.

The leaked info contains a few more specifics about the hand waving from a few months ago, but no actual evidence. What's worse is how the wording of the leaked document and press coverage (including that of The Intercept) interleaves simple assertions of fact into the narrative which do not support claims made by the narrative but simply make the whole thing seem "fact-based" even if it's speculative.

Generally speaking, The Intercept publishes very high quality work, but the presence of Sam Biddle's name on the byline is generally a good indicator of low quality work.

The difference between these cases is that even considering everything that is wrong with US administration, it is just not comparable to the Russian. People (and institutions) have a track record, and there's no principle in logic that prevents me from using that record to evaluate their trustworthiness: What you call an ad hominem, I call bayesian reasoning.

For example: It was, until this administration, extremely rare for the US government (the executive, to be precise) to lie to US-based press. They'd deny to comment, or say something that was meaningless when examined closely, or tell you to ask X (who will deny to comment). But, contrary to common believe, it's extremely hard to find examples for them straight-up saying A when knowing that it's really B.

It's evidence of what the person who wrote it thought at the time they wrote it. Your epistemology seems self-serving.

This is not a PR document, it is a (leaked) internal classified NSA intelligence report. It was presumably prepared by NSA intelligence analysts, who are experts in analyzing intelligence.