Returning 200 OK does not guarantee webhook processing

A surprising number of systems acknowledge webhook requests immediately but still lose events later during retries, queue failures or downstream outages.

While building FastHook I realized that "200 OK" often only means the request was accepted — not that the event was actually processed reliably.

This is why replay systems, durable queues, retries and observability matter so much in production webhook infrastructure.

Curious how other teams handle webhook reliability at scale.