You changed the scanning interval from 5min to 1min on these sensors.
PRTG ignores values in the calculation that are too far apart. The formular for this distance is 3*interval + 2*timeout. The idea is that if the scanned values are further apart it is likely there was a problem and the values could cause a false calculation. Like overflows that can not be relieably detected in this case.
With reducing the scanning interval this distance changed from 910s to 190s. Since you were monitoring with a 300s interval all values are ignored.
You can get PRTG to use the values with setting a high SNMP timeout value (like 90s). If you have frequent timeouts in SNMP you should reduce the value again after the report to prevent a lot of sensor hanging in the timeout.