Unpopular Opinion: "99.9% Uptime" is a vanity metric.

We obsess over "Server Uptime." We pay for Status Pages to show off our "99.9% Green Bars."

But let's be honest:
How many of us actually monitor "Functional Uptime"?

I audited a client last week.

Server Uptime: 100%

They lost leads for 12 hours because their monitoring tool was "Happy."

The Question:
What stack are you guys using to catch Logic Failures (not just server crashes)?
Playwright? Cypress? Or just waiting for user reports?