Anyway, unlike some, I find the extra "reach" to be real and advantageous.
Got any comparison images to back that up Bob?
Interesting debate. I go back and forth on this.
Since buying a 5DIII about a year ago, I admit my 7D has pretty much been sitting, gathering dust (although I keep it as a back up). But, in the past year, I've had very little time to shoot distance-limited subjects as demand for portraits seems to take up most of my spare time these days.
But, I can't quite understand how a crop sensor would never provide an advantage in distance-limited situations. I'm certainly willing to agree that if you crop the full frame down to the same framing as a uncropped APS-C image, you won't lose much, if anything. But, intellectually, I can't get my head around the idea that if I need to crop the image much more significantly, having the extra pixels of a crop would not be an advantage.
Being math-challenged doesn't help, but it certainly seems from a logical point-of-view that eventually, as you slice and dice away pixels, you'll reach a point where the full frame image loses too much resolution and you'll be better off with the crop sensor's greater pixel density.
It might require some pretty radical cropping, but then again, I've been in situations that require radical cropping (A California Condor perched on the top of an outcrop at the Grand Canyon – absent the ability to fly, you can't get any closer than the edge of the Canyon.)
There are several points to make, first is I am a little maths challenged too and agree, it doesn't make sense, but my empirical results illustrate my point.
Secondly, I never said it will never make a difference, but in my empirical testing (on older generation bodies but same theory) I found it didn't make enough of a difference to be noticeable in big prints even in optimal conditions set up to favour the crop camera. I would love Bob or Lee Jay to actually come along with some decent examples illustrating their beliefs, I have done so for mine. I am not saying 'I am right you are all wrong', I am saying 'I found this to be true, has anybody doing the same comparison found the same?' and several, like Neuro, have. I have never seen anybody post direct comparison images that illustrate a very different result, certainly after optimal processing at any kind of actual reproduction size there never seems to be a difference.
Third, as I have always said, the feature set of a crop camera might well make it a better camera anyway, things like AF, fps, cost etc can't be ignored.
Fourth, my crops are actually set up to favour the crop camera in iso and aperture etc, but also I upsized the ff file to match the crop file pixel for pixel to 'level the playing field' for comparison purposes, this should favour the crop camera even more.
But in the end I base my buying decisions on empirical results, I can't see $1,800 worth of difference in those >100% crops to warrant the expense, and I was happy to pay $3,750 for a 300 f2.8 over a $1,455 300 f4.