I use epoxy or even sumo. Had too many failures with CA starvation in the tubes.
But when I did use CA for glueing tubes I would wait about 30 min. and trim the wood flush with the tube then reapply ca to the trimed ends to help keep the tube in place. Retrim and put it on the lathe.
CA will be more reliable if allowed to set over night.
If you are giving a class in penmaking, you have a cup of coffee, then hit with accelerator--those are the extremes. I have made many pens both ways. The class pens are more likely to fail, but then, there are other reasons for that, too.
About as long as it takes me to pick up the drill with the end mill. Little to no stress on the tube if the wood is being trimmed and stop when brass brightens. If wood is punky, I add a little thin CA to the end grain, and thin to any crack after adding in fine wood dust.
Remember that the accellerator doesn't cure all the way down the tube. You will only firm up about 1/16" of glue at the ends of the tube, so it's possible for the center of the tube not to have bonded to the blank if you crank up your mill immediately.
I have very little problem with the mill blowing out blanks. The more frequent failure is to blow up a blank on the lathe due to inadequate bonding of the tube to the blank. I think this is worse with soft or burl blanks. I've taken to pouring a little thick CA into the hole before I insert the tube. I stand the blank up and let the glue run down most of the way through the hole, then glue up the tube and insert it while twisting. I do waste more CA this way, but I think it makes a better bond on the tube.
I prefer overnight for anything but since using CA Glue means I am in a hurry I will come back after a couple of hours in some cases. I have gotten away with just a few minutes but would not want to do that consistently for many of the reasons mentioned above. sooner or later frustrating things are bound to happen.
As has been mentioned, milling puts little to no torque on the tubes. Therefore, I'm not of the belief that it matters how well the glue has set up in the middle of the tube for milling purposes. I tend to glue up ten to twenty blanks at a time, hit them with accelerant, and mill them right away.
The only time that I have failures is when I am overaggressively milling. These failures typically happen early in the milling process and are not believed to be glue-related.