<quote>For Science articles, what value does all this narration add? Why do we have to sift through 30 years of background biography material before getting to the new theory and subject matter? Is the state of mind of the researchers important? Do we need to humanize and narrate the proposal of a new model in this way?</quote>
That's what popular science sites/books <em>do</em>. Maybe you should be seeking out the original research article.
Controlling the exit node does not inherently compromise anonymity, only allows access to the plaintext message. If using end-to-end encryption on top of Tor, you should be able to maintain both anonymity and protection from tampering and snooping.
They still don't know the source of the communication. Running an exit node means they can see where traffic is headed as it emerges from Tor, not that they can magically determine from whence it came.
The problem here is that the destination can identify you. You, after all, are the one paying for that VM instance at that IP address within AWS. Who else would be connecting via SSH?
That's what popular science sites/books <em>do</em>. Maybe you should be seeking out the original research article.