Sony shows specs on some of their high end cameras as having one resolution
for 16x9 (usually around 700 lines horizontal) and another, higher, spec for
the same camera in 4x3 mode.

I cannot for the life of me understand how this can be possible given the fact they are cutting off pixels on both sides of the CCDs in 4x3 mode. If anything, it would have to be lower. Less pixels = lower resolution, not higher, correct?

Can anyone explain this?

I've never seen those specs, but here's what I think is going on. 16:9 DV is anamorphic which means the same number of pixels are being stretched wider. So 720x480 is the resolution for 4:3 and it's also the resolution for 16:9, but in 16:9 mode you're stretching the same information over a wider horizontal area which would be the equivalent of 960x480 (also remember that DV pixels are not square).

So regardless of the camera's CCD resolution, only 720x480 is written to tape if you're recording 16:9 DV.

The specs do not take into account the fact that the tape format is DV. I'm speaking of camera resolution, only.

I quote from a Sony camera sheet "The 700 lines of horizontal resolution at the 16:9 aspect ratio are actually equivalent to 930 lines, which are converted by the 4:3 aspect ratio camera measurement"

DV25 is limited to approximately 500 lines of H resolution, period.

"Camera lines" are measured in relation to image height, so the figures are smaller for the widescreen anamorphic format. It is an oddity in how that particular spec is measured, and shouldn't reflect negatively on the camera. If you are comparing two cameras, just make sure you are comparing "apples to apples," 16:9 to 16:9 or 4:3 to 4:3. There was a discussion of this recently on dvxuser.com. Search for "camera lines."