Most users will be able to answer “yes” or “no” to “Do I want to see this at the top of the page?” about most comments, I think. In which case, if they follow CS’s two rules then they will vote those comments either up or down: so most comments end up getting voted on by most users.
Therefore, any comment for which a fraction of users substantially less than 50% wants to see it at the top of the page—and, as you rightly say, that’s likely to be most of them—will get downvoted heavily if everyone follows CS’s rules; almost certainly to lower than −4, unless voters notice soon enough that this is happening and switch to following Eliezer’s rule (which they cannot do at the same time as following CS’s rule).
(I’m sure CS didn’t intend his rules to be applied so mechanically, and actually I think both his and Eliezer’s are basically correct.)
Most users will be able to answer “yes” or “no” to “Do I want to see this at the top of the page?” about most comments, I think. In which case, if they follow CS’s two rules then they will vote those comments either up or down: so most comments end up getting voted on by most users.
Therefore, any comment for which a fraction of users substantially less than 50% wants to see it at the top of the page—and, as you rightly say, that’s likely to be most of them—will get downvoted heavily if everyone follows CS’s rules; almost certainly to lower than −4, unless voters notice soon enough that this is happening and switch to following Eliezer’s rule (which they cannot do at the same time as following CS’s rule).
(I’m sure CS didn’t intend his rules to be applied so mechanically, and actually I think both his and Eliezer’s are basically correct.)