Second, it will be included in the Linux kernel starting with version 3.5 according to that page. Whether dd-wrt can take advantage of it remains to be seen(drivers) and it also remains to be seen if it will even be implemented at all.
Remember, all of this is in the experimental stage for now. I personally only do the txqueuelen method to combat the problem(if it even exists) by setting it to 100 instead of 1000 but that's just me.
Sorry if I got the spelling wrong. The name of the podcast episode seemed to indicate that it was CODL. According to Steve Gibson (whom I trust) it's a true solution to the problem of buffer bloat. I don't like upgrading DD-WRT often, but this would definitely be something worthwhile to upgrade for, so I figured I'd ask. Sounds like the algorithm is pretty simple to implement, so it seems very technically feasible to implement in DD-WRT.
Wow that was quick. Well I'm hoping that it can be implemented and used in a future build of DD-WRT. If and when that happens it should be clear in the build notes because this is a significant feature. Anyway thanks for the update.