When you run a report on one monitor for February, the times are correct.
When you run a report on a group of monitors (e.g. two monitors) the time goes from 28 days to 56 days.

This is not logical. February only had 28 days, not 56 so uptime should be 28 days if it was 100%, not 56 days.
It seems that the logic for calculation of the times on groups of monitors is too simple and makes no sense if you have multiple monitors in a group.
The time just gets multiplied by the number of monitors...