The effort may outweigh the costs if there are higher value projects this would be prioritized against (including the on-going support burden)

For example, a project that made Chrome users 1% better off is probably worth far more to the overall 1pw user base.