I have used a ResMed Aircurve 10 machine for a number of years now and was unaware of the existence of this software. It appears that the software has to be purchased under the advice of a physician. Not sure why any patient would want or need it. The important things are there for review with MyAir as you noted. AHI is the most important statistic by far combined with hours of use. The ability to control apnea and the hours a person sleeps seem most important. Monitoring mask leaks can contribute to finding the right mask and that helps with apnea mitigation and the ability to sleep. Also monitoring mask on/off will be helpful if a person is prone to taking off their mask. The MyAir software doesn't monitor humidity level. Single most important thing is AHI as that is a measure of the effectiveness of therapy but the statistics contribute to the effectiveness of therapy.
I use ResScan all the time (and hate it because it is unstable and unreliable, but that's neither here nor there). As Wiredgeorge said, the main statistic is the AHI. This measures the number of changes in breathing that occurred per hour of operation (not sleep mind you because the machine doesn't know when you are sleeping). Generally speaking, anything 5 or below is good. The hours of use is also important. The most useful figure is average usage per night (this is what is used for assessing compliance). ResScan also supplies median usage, which is a bit obscure and not terribly useful. Mask leak can also be handy, which ResScan can show as median leak, 95% leak and max leak. median leak is similar to the average level and should probably be below about 15, but this is rather variable depending on the patient and the mask used. Max leak level is a bit useless because it is the maximum leak level recorded at any point in the period of the report, which may only last for 2 seconds and is therefore not relevant. 95% leak is more useful because it is the highest leak rate reached if you leave out the worst 5%. This is handy because, combined with the median leak rate, you can tell if fit is usually good but gets worse in patches or if it is good all the time.
ResScan is highly configurable and can display all sorts of statistics and graphs. Most of it is totally useless. Stick to the basics. Also remember that the data should always be taken in the context of the patient's real world experiences. If there is a high leak rate but the patient is happy, the AHI is low and their symptoms are well controlled, why bother about it? Likewise, just because the leak reading is great that doesn't mean you can simply ignore the patient telling you that their mask leak is bothering them.
As usual, Wiredgeorge is on the ball.
I do think that if a patient is new to PAP therapy and received their equipment via insurance, hours of use is kind of important because the insurance company also finds out that statistic and if a patient is NON COMPLIANT (usually under 4 hours of use per night), the machine is taken away. I am a member of other forums where folks post all kinds of graphs and facts/figures from 3rd party software and I just find too much info makes my head spin and in my case, at least, doesn't contribute; only confuses unless you have a professional looking over your shoulder to interpret.