Floydo, based upon tests I ran at the time 2012.20 was released (see [Only registered and activated users can see links. ]), there is a significant reduction (estimated at 90%) in the number of junctions with images in the small file, regardless of whether the images in a JCV file are photorealistic or not.

While I don't agree with your testing method, I agree with the general sentiment in that Frustrated should use the large (750+MB) file since it has all the junctions and displays them in full-screen glory. A proper test should not focus on just one intersection, but should run a route, such as what I did for my tests.

Now, I am only speculating on this, but I believe we should have six different JCV files:
  • Small JCV file for legacy units supporting full-screen Junction View (anything pre-2012)
  • Large JCV file for legacy units supporting full-screen Junction View (the 750+MB file for anything pre-2012)
  • nüvi 30 JCV file (non-photoreal full-screen views)
  • JCV file for nüvi 40/50 and legacy units supporting split-screen Junction View (photoreal split-screen views)
  • JCV file for Advanced units (photoreal split-screen views)
  • JCV file for Prestige units (photoreal Bird's Eye split-screen views)

I'm sure we'll see more files than this since Garmin likes to screw things up.