Why isn't mutation testing widely used?

Testing and TDD seems to be very widely used, with people disparaging anyone who doesn't write comprehensive tests.

However, most organizations still measure test coverage using branch/line coverage metrics. This seems heavily flawed; it's extremely easy to write tests that exercise all code paths, but don't actually catch bugs in those code paths.

Mutation testing doesn't have this problem, and seems far superior to line/branch coverage metrics in measuring test suite comprehensiveness. And yet, it's seldom even referred to. I see TDD, unit-tests and code-coverage-metrics referred to 100x more often than mutation-testing. On stack-overflow, code coverage has 2000+ tags and mutation testing has just 40.

Why isn't mutation testing more widely used?