There are a couple of Cribl BUG that are contributing to this behaviour.

CRIBL-18868:
The monitoring dashboard for destinations is showing the health from the earliest timestamp of the range, instead of the latest. This means, with a 1h time selection, the health status is delayed by an hour.

CRIBL-12259:
The Health metrics were disabled due to a performance issue with the GUI when the numbers of workers exceed 50+ workers as per CRIBL-17756 and on the destination page we default to the warning status.

Thanks! I look forward to the fixes!

CRIBL-18868: Has been fixed in v4.4.2

CRIBL-12259: Has been fixed in v4.3.0