jbu-
Wrong.
The F stop ("2.8")has nothing to do with the parallax error. All that an F-stop referrs to is the variable aperature within the lense which controls the ammount of light transmitted.
All the use of a lower F-stop lense the would have allowed the shooter to do is eliminate any pin-cusion effect at the corners by re-locating the camera further back from the subject and then zooming in while maintaing a large apeture, which in turn would have admitted more light to the image surface, allowing for a higher shutter speed.
The problem with this is that when running a large open aperature on a lense with a longer focal length is that you must sacrifice depth of field (the parts of the picture that will be in focus) for the additional transmitted light the large aperature will gain you.
An additional problem that plauges all telephoto and zoom lenses is depth compression that occurs at the longer focal lengths. This phenomina will cause the subject to look flat. The longer the lense, the more severe the compression.
The paralax (convergance) error is created by tilting the camera upward. Paralax can only be eliminated by keeping the image plane both perfectly parallel and squared up to the plane of the subject matter. The only way this can be accomplished is with the use of a tilt-shift lense.
What a tilt-shift lense does is allow the photographer to capture a similar subject area without tilting the camera by introducing a set of independant movements into the lense element. Light transmission is achieved via the use of flexible accordian-like bellows between the lense element group and the camera body.
In this photograph, the camera would have been squarred against the structure on the left side of the building then locked in place, while the lense would have been shifted up and to the right.