I am very grateful for the help I received from several people (mostly off list, which is great to avoid spamming the list).
Let me also use this email to briefly comment on two points from Matthew Walster's posts; and Matthew, I really come at peace, I have a lot of respect for you and your work, but we can also disagree on some things, right? So:
1. Matthew's email basically seemed to imply intentional hijacks are not a concern (rare/non-existent?). Few measurement works seem to show the contrary; I esp. recommend the `Profiling BGP serial hijackers' paper from IMC'19 by a team of excellent researchers.
2. A bit off-topic, Matthew's response to Dora Crisan seem to imply BGP eavesdropping for eventual cryptanalysis, possibly using Quantum computing, isn't a concern. On the one hand, I agree that Quantum computing seems still quite far from ability to break state-of-art PKC, and it may long till it becomes practical (if ever). OTOH, it may also not take that long; also, `conventional' cryptanalysis may still happen, e.g., see Schnorr's recent paper,
ia.cr/2021/232, which claimed to `destroy' RSA [withdrawn later, so apparently even Schnorr can err - that's part of science - but this doesn't mean next effort won't succeed or that some TLA (three lettered adversaries) didn't succeed already]. TLAs may have other motivations for eavesdropping, like collecting meta-data. Now, I am sure many customers and providers may not care about security against such TLAs, but I think it is legitimate for some people to be concerned.
Best, Amir
--
Amir Herzberg
Comcast professor of Security Innovations, Computer Science and Engineering, University of Connecticut