Oh, by the way, I didn't call myself 'Gamer' because that (searching pics for all vehicles) was my 'game', before you were going to say that.

I could make that from two text lists indeed, that would be easier

Actually on the server images are stored in a folder from the first 3 digits of the number, to avoid having more than 1000 pics in the same folder (or else it takes an awful long time to access remotely).
Making a list of files across all subfolders is still easy, though that a list of 600000+ files takes several seconds to make - so I can't really make a dynamic page listing that in public.
So that would be a simple static list of the IDs. But I guess that for Gamer's game of completing missing vehicle pics it would already be enough.

I'll do that when I've some time in the next days.

Isn't every vehicle main pic named i[id#].jpg? Can you just diff a list of all those image files against a list of all vehicles?

There no picture info in the database so it is not possible (well, technically there could be a way of getting the list, but as the site as more than 600 000 vehicle entries, it would be way to heavy to do it).
I guess it was handled differently in WebCarStory, many things were changed in the different sites since the fork of the code.
In the next version of IMCDb that will be possible.

In my opinion, we should have an overview of all vehicles lacking pics such as: http://www.imcdb.org/vehicle_104594-Volkswagen-Golf.html

Besides, Web Car Story also has that feature.