Most of the graphs we publish that show climate projections are based on simulations that are run as part of the IPCC assessment report process. So, we ourselves are not running climate simulations and choosing the end point. My guess is that the reason they don't usually project out beyond beyond 2100 is that it takes massive amounts of computer processing to run these experiments, and they have to choose a reasonable cut off to keep the processing manageable, but I am just guessing. It may also be that the range of uncertainty gets so large as to make the projections not useful. However, I do know that i have seen a few graphs in the IPCC physical science reports that go out farther than 2100. I would check out Chapter 12 (long term change) in the IPCC Fifth Assessment Report, accessible here: http://www.ipcc.ch/report/ar5/wg1/