We plan to quickly iterate on this approach a bit (as well as extend it to posts created via the API). So far, we have observed a slight decrease in the amount spam being created. There's clearly more to do.
Update: Some users are having trouble solving the new CAPTCHAs and are seeing CAPTCHAs when they shouldn't be. We'll be pushing out a fix for that shortly.
Update: This should now be resolved.
