Sounds like a pretty good experiment. If you are running a carbon reactor, you could probably omit it and run carbon similar to how I am running (floss-carbon-floss sandwich). It seems to be working quite well to force the filter floss to capture more precip.
That is, if you needed some space savings. I also feel that since you are going to run sand, you will likely catch most of the stuff in that chamber anyway.
My only concern about such a small sand filter is that you may have problems getting a proper flow rate. Too slow, and you wont be moving too much water through it due to quick build-up. Too fast, and you will just blow all the precip past the sand and into the 3rd chamber. I prefer the slowest possible flow rate (within reason). I want to give the LC the longest possible contact time.
I'd love to hear how it works. I'm still very happy with mine. Time will tell all!