Hi all,
there is a discussion in T6845 about experimentally disabling the captcha for a short period of time and seeing what effect that has. We have long known that our captcha can be trivially broken by off-the-shelf tools and such does not provide any barrier to a minimally dedicated spammer (while it does provide a significant barrier to users with disabilities and non-English speakers). That's the theory, any way; it would be nice to get some empirical evidence.
Your feedback on this plan would be appreciated.