[response requested by Sept. 7] COGA response to WCAG 2.2 issue 1902, common objects

Hello COGA task force,

*Summary: *
Please review https://github.com/w3c/wcag/issues/1902#issuecomment-907469906
and follow up with your concerns or comments.

I will attempt to draft a revised proposal based on your feedback on
Tuesday, September 7.

*More details on this request:*
As discussed in our meeting today
<https://www.w3.org/2021/09/02-coga-minutes.html#t02>, we've been trying to
find the right response to WCAG 2.2 issue 1902
<https://github.com/w3c/wcag/issues/1902>, which started around common
objects but has become significantly larger, and is now more of an issue of
including cognitive function tests in the authentication process.

We have engaged in a long process of working with numerous parties on
this issue, including:

   - Abi James and John Rochford worked separately with Alastair to attempt
   to resolve this. The proposed resolution was rejected by COGA as
   insufficient.
   - Alastair and Rachael have been helping us to try to understand the
   legal and technological constraints that will block consensus.
   - I personally spent substantial time going over this with colleagues
   who are working on Google's CAPTCHA 3 and who are are responsible for
   safety, security and privacy.

Ultimately, I attempted this wording, which I'd like for everyone with a
stake in this to review and comment on:
https://github.com/w3c/wcag/issues/1902#issuecomment-907469906

There are 2 options presented here. Option 1 likely won't gain consensus
but I believe better reflects COGAs perspective. This is firm language that
essentially rules out captchas entirely. Option 2 offers a compromise of an
AA/AAA split, and I believe that this option would be able to get consensus.

This approach has some linguistic elements to clarify, and also still
requires Lisa's addition of making it clear that if we go with option 2, an
AA / AAA split, we need to add language that emphasizes that some will
still be excluded.

Thank you,

Rain