It is not that 1 on 4 flights were delayed, just that 1 in 4 were not on time. This could mean early as well. I myself have been on international flights below the equator that arrived ahead of schedule.
And this 75% was for April 2011 only. It is quite possible that other months the on time rate was different, even 90%. What you would want to do, is find data for on-time rates across a wide range of years, not a single month. Then we could show that above the equator, flights on average arrive ahead of schedule, whereas below the equator, flights arrive too late.
This would help support FE. Sadly, the linked article does not.
You may also consider planes departing late.

The arrival delay is the sum of the delays, not only the travel time one...
To be fair, you should also take a look at the return trip, assuming that the flight paths are identical (they are often different)
I've found a useful tool to gather informations :
http://www.flightstats.com/go/FlightStatus/flightStatusByAirport.doPlease select "On-time performance ratings" item.
You can get ratios and more interesting, mean delay times. You should note for instance, for a LHR-JFK travel and the two last months, that the best result is worth 80%.
In fact the real interesting value should be the delay between crew-planned flight times and the actual ones. Unfortunately this is not a public information.
This raises another previously raised question: As the fuel is one of the most watched parameters on a plane,how such alleged discrepancies are unnoticed every day on every single flight?