…even though both of them are representations of Layer-3/similar machine-readable addresses*

However, there is such a thing as "inviting people to beat you up in such a way as to draw media criticism without plausible likelihood for constructive input", and I feel that this would be onesuch.

Experiential evidence:

1) the number of people who've told me in-past that Email addresses are unusably unmemorable, except somehow 30..40 years later we are still using them, and have developed coping strategies, eg: address books.

2) the number of people who've told me in-past that IPv4 addresses are unusably unmemorable, except for 8.8.8.8 and 192.168.1.1 which somehow are enough for people to bootstrap access to the rest of the internet, and use various coping strategies (eg: DNS, bookmarks)

3) the number of people who've told me in-past that Old-Style Onion addresses are unusably unmemorable, until (as mentioned above) Facebook and a few other good ones got mined, and people started taking Onion networking mildly seriously as a means of more-secure enterprise communication… Oh, and bookmarks as a coping strategy.

4) phone numbers. unusably unmemorable. coping strategies: in-phone address books + address-book synchronisation. etc etc etc.

So: can we do better with Onion UX? Certainly.  

Should we research improvements to user experience? Absolutely.

Should Tor invite opinionated people to come piss all over its equivalent of https://[2001:0db8:85a3:0000:0000:8a2e:0370:7334]/foo.html? Probably not. Just my opinion. I don't feel it would benefit anyone except (a) haters, and (b) academics who research only "what doesn't work" because researching "what /does/ work" is beyond the scope of their funding.

     -a

* explanatory thread: https://twitter.com/AlecMuffett/status/802161730591793152