1. With rates per 100 departures rather than just a static number - 16% of 1000 departures might be the same performance as 0.2% of 2 departures when it comes to the actual number of passengers inconvenienced.
2. Breakdown of what caused the cancellations; maintenance, spare parts, crews and a lack thereof, weather - one mid winter storm here in the Midwest would easily decimate NWA's cancellation performance for a month.
3. Aircraft types - one fleet may have a more restrictive MEL at one airline than another, and thus a cancellation for one type is a departure for another. For example the cancellation rate for NWA A320's was always higher than the cancellation rate for the same Air Canada fleet because NWA's MEL's were far more restrictive.
4. Most useful is whether this is a one month phenomenon or a trend. One set of data doesn't really give you a broader view of comparitive performance.
5. These data are basically arse wipe.
I definitely appreciate your knowledge, understanding and deep analyzing skills of Airline Operations but one thing uve missed out is you are talking about Go Air here and let me share with you my experience with go air (this happened before I came into aviation)4 years back .. I had booked an Mumbai - Bangalore ticket for a friday and on wednesday evening I receive an information(SMS) that the flight has been cancelled due to techinical snag
!