How should I handle the following disagreement in a code review?

I am working for a fintech startup where we are heavy on unit tests. For a particular release, I spent an entire day writing various unit tests (over 100). My colleague, who is peer-reviewing my code, has suggested that I should write some more tests for some cases that are extremely unlikely to happen. You can say it's extremely rare for those cases to occur. The current test suite covers 99% of the commonly-occurring scenarios.

I admire my co-worker, but I don't agree with this particular approach. What do you think? Should I spend time writing more unit tests for cases that are extremely unlikely to happen? This will delay the release for another day.