Latency is showing incorrectly on homepage

[ckbedwell]

The "Home" dashboard shows traceroute and scripted checks as "red" even when they're performing excellently. My guess is the "latency" on the home page is based on runtime (with a delay on K6) rather than the actual test latency. The result is the "red" in the latency column looks like trace and K6 tests are failing, when in fact the targets are performing excellently.
On the home page it shows 2+ seconds. Yet, on the drill-down results for this check, the latency value shows 198ms.

Slack thread for context.

[peterschretlen]

The homepage looks at probe_duration_seconds to generate a table (all check use the same metric to generate this column)

On the check dashboard, a different metric is being used that accounts for the overhead of k6 startup