There are actually some insanely incredible things happening using the .NET Framework. One is a full implementation of the distribution / remoting / analysis software for the huge particle accelerator out in Geneva which'll be up sometime around 2010.
I know a few people in various universities who were sick of the awful implementations of the existing software and are rewriting these insane apps in C# using .NET.
Very, very cool.
I'm very glad to hear this. Researchers need to spend most of their time thinking about specific problems that are directly related to their research, not problems that arise from the software tools used in making their research more effective.