Big skimmers are a myth when it comes to bio-pellets. Any skimmer will do and they only work for a short time. AdvancedAquarist proved that.
http://www.advancedaquarist.com/2011/3/aafeature
Which leads to the question: Is the bacteria that is skimmable good or bad? If good, wouldn't you want to leave it in the tank? Is the bacteria left behind good or bad?
First time I ran pellets for 6 months.. crash happened at 2 months but I 'stuck with it' to see if the tank would recover, by month 6 I had a lovely tank of white sticks. Second time I went 30 days and had several corals die. In the 'modded' TLF reactor with a maxijet 1200 the pellets would clog in under 24 hours, in the BRS reactor they would clog in about 3 days run off my return pump. First attempt I used 1/2 recommended dosage, second attempt I used 1/4 recommended dosage. Both times I used pellets caused my ATS screen to turn to mush also. I came to the same conclusion as the guy in the video, without a way to control the bacteria populations, it isn't a viable method, and up until the recent recirc reactors you had to push a ton of flow through them just to keep them suspended.
I am not against carbon dosing in general, I actually had great success when I dosed vodka (and I believe it was because I had control over the bacteria populations because it was dependent on my daily liquid carbon dosing), I just don't think the current method that most people use today for using pellets is correct.
Ya Garf, the solution is in the video above.
Use one of those and just have the flow so slow it is just a steady drip in the sump and it may work good with an ATS. The key is to be able to test the water and adjust the flow so you do not 'bottom out' the N/P in the tank. If you can control that (which you couldn't with old bio-pellet reactors), you will be good.