I don't think you would if your upfront goal was "telemetry" and you actually evaluated solutions for that. But if you had the Sentry pipeline in place already for crash reports, it would be very easy to slide into using it for tracking other kinds of custom events.