From page 10 of the 10-Q: "During the second quarter of 2009, the Company reassessed the allocation of total premium costs between launch and on-orbit insurance and the expected consumption pattern of the total insurance premium costs. Based on its change in accounting estimate, the Company reclassified approximately $22.0 million from prepaid insurance expense to the capitalized costs of the GeoEye-1 satellite. As a result of this change in estimate, net income before tax and diluted earnings per share for the quarter ended June 30, 2009 increased $9.1 million and $0.28 respectively."
So instead of the reported 46c of earnings, they actually did only 18c! This would have missed the consensus estimate of 39c by a wide margin. Also means that the actual earnings for the year are below $1.
As they said after the Enron collapse: people would have known it was a scam had they read the footnotes....
The quote from the 10-Q is, in fact, accurate. However, this new treatment of the insurance premium was mentioned in last quarter's CC. Therefore, the consensus estimate likely took this into consideration. In which case, it's still a beat.
In effect, the 10-Q statement said that the new treatment added $0.28 to income relative to previous quarters. It would be incorrect to interpret this to mean that $0.28 of income was created thru an accounting trick.