In the beginning I found myself in a similar situation, and started looking at the 2 minute rule as more of a guideline, as Mr. Allen points out:
Two minutes is in fact just a guideline. If you have a long open window of time in which to process your in-basket, you can extend the cutoff for each item to five or ten minutes.
He also points out: Process does not mean "spend time on."
So, it kind of depends on your circumstances and your better judgment. If you have been inundated with a lot of things in the in-basket and you don't really have time to do much other than process them, you may find it helpful to not actually do the two minute items, but put them into a pile/list all their own (or, just add some visual indicator that says, "Won't take too too long"); or, as Mr. Allen points out, shorten the time - if you think it will take one minute to do - go ahead and do it. However, if you're sitting around on an otherwise idle Tuesday and have some real time to kill before you're next calendar item - go ahead and finish. Finally, if you reassess and realize this item will take you hours - get a stopping point, add a new action, and move on to the next thing.
I've found doing this usually takes me to a pretty happy place quickly. You may also want to examine how good you are at estimating in the first place. Take every action and mark how long you think it will take to complete. For those things that will take multiple days - break them down as a project into finer and finer levels (this really helps me). Just add the items as you think of them, estimate how long it will take, if you say more than a day - you don't have to stop and try to break it down now - but, your mind will probably start thinking about it without you being consciously involved.
Hope that answers the question. If you would like me to expand on any of these, let me know.