I want to be able to see if we've improved or regressed in code-coverage between two different commits.
So what's important here is not just the coverage, but understanding what lines of source should be part of the report, and which should be ignored.
As I maintain a project, and I get a Pull Request, I want to see that the person has not regressed in code-coverage on the existing code base. I have the two commit IDs. I can run dotCover against each and then produce the report.