I'm afraid you don't have that quite right. "Bad" triple buffering without v-sync still adds input lag and at the same time allows tearing. The point of triple buffering (well, "good" TB at least) is that you can run a game with v-sync without stalling the GPU due to a full back buffer, and with lower average input lag than double buffering with v-sync. If you turn off v-sync, you've completely defanged triple buffering - why have a 3rd buffer if you're just going to immediately buffer swap anyhow?
On 60Hz monitors triple buffering would add another 16ms of input lag, which is not bad for most single-player games (or any online games that do not rely on as little "ping" as possible).
Usually, turning off vsync means that there's no triple buffering (not even the "bad" triple buffering), even in many DX10/11 games that automatically call for triple buffer whenever vsync is enabled. In the old days, many games would offer TB as an option in the graphics settings.
I would say that adaptive vsync benefits the SLI gamers the most since Alternate Frame Rendering usually does not properly do true triple buffering anyways. AFR already calls for 4 buffers (imagine the input lag) with simple vsync, and true triple buffering does not work as seamlessly in many games. For example, in Source games that offer TB as an in-game option, it would be "pseudo" TB where the frame rate drops from 85 fps to 60, then to 45, then 30.. a very strange phenomenon that is not really any better than double-buffered vsync.
Even though double-buffered vsync theoretically adds only between 0-16ms of lag to a 60Hz screen compared against disabled vsync (in which some areas of the screen is updated faster than the tearing part which is updated up to 16ms later), vsync itself usually detracts from the "ideal" responsiveness of input devices (especially the mouse) - which is rather perceivable in the majority of games out there. To see if it helps, try forcing the number of max prerendered frames in the CP from default 3 to 2 or 1 for better responsiveness with vsync. If the setting of 0 still does not improve the responsiveness of your mouse, then it's probably better to just disable vsync altogether if the tearing is not too unbearable (which unfortunately usually is the case with 60Hz panels).
On my Sony GDM-FW900 which can do all kinds of refresh rates, even overclocked to 2560x1600 @ 69 Hz, I play many online fps shooters at 1920x1200 @ 90Hz. As long as the frame rate in the game is also 90fps most of the time (EDIT: manually locking it at 90 fps), the tearing is rather minimal. The output seems to already be "synchronized" fairly well in that it does not really need vysnc, and the responsiveness is as good as it gets without having to deal with the input lag and response times of LCD monitors.