Consider the following two households:
The Johnstons is a family of 5 (Mother, Father, teenage daughter, with two sons [4 and 11]) who own a Wii which receives an average of 50 hours of playtime a week (Mother - 5 hours, Father - 7 hours, Teenage Daughter - 8 hours, and each son plays for 15 hours)
The Smiths are a family of 5 (Mother, Father, teenage daughter, with two sons [4 and 11]) who own a PS3 which receives an average of 40 hours of playtime a week (Mother - 0 hours, Father - 10 hours, Teenage Daughter - 0 hours, and each son plays for 15 hours)
How would this data be interpreted by the Nielsen survey? Depending on the methodology taken you could have completely opposite interpretations of the Data ... On top of that, there is a great deal of difficulty accurately collecting data about users of a system which is more likely to have multiple users in a household based on a survey, because the person you talk to may be able to accurately report their own usage but is unlikely to report someone else's use accurately.
The fact is that the Wii sells a similar ammount of software per system sold as either other console, so the "Lower" play-time per user of the system has to be roughly balanced out by an increase in the number of users per system.
