I was running some PLA this last week and just fell in love with the stuff - it's basically as well behaved as acrylates are. The time-temperature superposition is very easy to run with both G' and G" coming together beautifully. I also get a good amount of horizontal shifting too, so that the master curve really takes in a wide range of frequencies. Basically, I look like a genius because the data is so good.
The other extreme that I have to face on a very regular basis is a PVC-wood flour composite made by our parent corporation. It's just a nightmare - got to move fast and can't go too high in temperature before it starts burning and degrading (both the PVC and the wood!). I end up with lots of scatter, a limited output range and data that is only maginally helpful to reach any conclusions.
I hate working with silicones as well because they are so insensitive to temperature that you really have to make large changes before you get anything to change.
So what are your dream and nightmare materials to test?