I would be especially interested in the color quality of the images. I'm trying to find a new camera for our histology teaching laboratory, and in this case we need live full HD image with good color reproduction.
I have limited experience, which really means I don't recall being so impressed by it that the camera's images clung in memory. I AM impressed, however, by Canon cameras, and I have run side by side tests against many scientific cameras (not including Leica) and find that the color is truer (though never as true as when a color calibration slide is used: there is no such thing as "true color" until matched to a standard). You can come closer to true colors when white balancing manually, true for all cameras. Canon cameras have HD capability.
You can see a comparison against one scientific camera (which isn't named) at http://imagingandanalysis.com/products_canon.html and a link to a person who not only sells them, but will send you a camera for testing.
If scientific cameras use mosaic chips, many use the Sony chip, which, in my imaging research, tends toward yellow, even after white balancing. However, scientific cameras often use algorithms that saturate colors, so you may be more interested in a scientific camera after all, especially when you prefer colors that appear more "neon." Many scientists do, so it's really your call in the end.
Thanks for the commments Gerald! I agree that Canon SLRs generally produce much better image quality than dedicated scientific microscope cameras, and we in fact just installed one in our cell culture lab. However, a "real" microscope camera with a dedicated software seems to be somewhat more fluent to use, which is very important when showing live microscopy directly to the students (= I don't want to spend an extra second for making the settings etc.).