Yes, they typically plot out to l=10000, but all the power out there comes from point sources (if one doesn't clean those away). I used 360°/l, but if I use 180°/l the ACT foreground cleaned spectrum becomes too faint to be detectable at 2.9 arcmin.