Then you can consider every quarter an anomaly.
You can say that when the 3G iphone was launched and Apple got a big market share --- it was because RIM hadn't launched a new phone model for a year. Since 3 out of the top 5 smart phones come from RIM, it's all about RIM and nothing to do with the iphone.
If we use your logic, then the 4 quarters look like this: July-Sept quarter iphone launches (massive numbers), Oct-Dec quarter (actually really bad quarter for the iphone because they managed to go flat or even go down in sales even with the busy christmas quarter), Jan-March quarter (have you visited a shopping mall during these months, it's dead for everybody), April-June quarter (Apple "ran out" of iphone inventory and have a token amount of market share).
Trends are more important than any one quarter overall.
But it depends on what happened in a particular quarter. As you say, we have to consider what was being released.
Also, is it a selling quarter, or a quiet one? Are most of a company's sales to business and government, or consumers? Who has come out with what? How far away is the next model, and how does that figure into declining sales for the quarter leading up to it?
I like to look at each quarters sales, and look into all of that. a single quarter can tell us something important, depending on the factors involved.
For example with Palm. Sales have been dropping pretty much continuously for a couple of years. Sales are less than a quarter (in dollars) of what they were.
I would expect Palms sales to be less because of the Pre. But as sales have dropped so much anyway, it's really difficult to determine how much of the past 6 months sales drops are due to the Pre, and how much is due to people just abandoning the platform.