The 24 hour test is one of the areas where status monitoring systems shine.
A common way to perform the 24 hour test is to use data logging signal level meters. The meters are left at the test site for the duration of test sequence (at least 18 hours).
Sending techs around the system with meters has the advantage that you don't need to worry about temperature effects on the SLM performance (plus you're pretty much assured that the meter won't disappear over night).
The 3 dB adjacent channel level requirement is only for channels "within a 6 MHz nominal frequency separation". Technically it doesn't apply to e.g. channel 4 to 5 or to other gaps within the spectrum (such as in the FM band).
If you plot only one part of your proof test data, the 24 hour data should be the one. It provides a good look at the performance of a large portion of the system. It also provides insight as to the results of other tests - such as C/N and In-Channel Response.