There were several problems with the script.
First, one monitor (itrzn) was computed much higher under radegund (Ubuntu, gcc v7.5) than on other testing platforms, which themselves have drifted since the monitored value was last updated (2014). Update the monitor to capture the current valid range with continuum resolution% error.
Second, about 30 monitors were not exercised because of an empty line that appears to have been accidentally inserted in 2009. Reenable and update monitors.