Has anyone grown faster by killing their most requested feature?

We spent almost three months building and maintaining a feature that 40% of our free users asked for. But when we looked at the data, that exact feature was causing 60% of our support tickets and confusing high-paying customers.

Last month, we ripped it out completely. We lost a few vocal free users, but our main product speed improved, support volume plummeted, and paid upgrades went up.

Have you ever sunsetted a popular feature to save your main core value proposition?