There should be no problem with running it in Moldflow. I've run a few micro parts before and it worked well. I don't remember trusting the injection pressure numbers, but the filling pattern made sense.
One challenge with micro-molding is understanding what is actually happening in reality. Typically the molding process is very fast, and the machines may not be responding as quickly as you anticipate. If you for instance specify an injection speed of 0.1 second, the machine will need some time to accellerate over a very short distance, and as a result the screw will probably not accellerate (and decellerate) very linearly. Measuring the actual screw speed over a short distance is not very trivial. Also the inclusion of venting in the mold can be important.
Within Moldflow we don't know the machine characteristics, so the assumptions are that they are 'ideal'. At the university of Danmark they did some work that was published in this paper (I believe there are more).
I hope this helps
Product Manager for Autodesk Simulation Moldflow products