A couple of thoughts:
One is that I'd like to know how things scale in the real world according to displacement. Intuitively, it would make sense that it would scale linearly. With that said, a lot of small econobox engines idle at 800-1000 rpms while it's not uncommon for a big American V8 to idle at 600-650rpms or sometimes even slower.
With all of that said, I always think about my little MG with or really any carbureted car. The fuel "consumption" of a carburetor on a hot, stopped engine is not zero as some gasoline will always evaporate out of the float bowl. My MG(and most cars made up until the very early 1970s) is completely "open" and evaporated gas vents into the atmosphere. There are a lot of interesting things at play in this sort of situation, as in operation the constant atomization of gasoline in the carburetor throat keeps the temperature of the carbs relatively cool. In operation, a typical set of numbers might be 200ºF on the cylinder head, 150º at the intake manifold, 130º on the float bowl, and 120º in the throat. As soon as the engine is shut off, the intake and carburetors start to heat soak and in a minute or two will equilibrate out to probably 190º or better all over. If I shut the engine off in traffic, on restart(even after a minute or so) the fuel pump ticks a few times, signifying that the float bowls are probably down milliliter or two from full. Even more importantly, though, you end up with fuel in the throat of the carb(not sure if it's just simple fuel expansion or percolation, or a combination of both but whatever the case it happens) so the car runs EXTREMELY rich for several seconds on a hot restart and I suspect emissions(particularly CO and VOCs) are through the roof.
Carbureted vehicles don't really even enter into the big picture these days as there are not that many being driven these days and on the whole the ones that are driven tend to not rack up a lot of miles. I imagine that of the ones still racking up significant miles, vehicles without evaporative loss control systems are an even smaller portion(although that system doesn't fix fuel percolation and rich starts). Still, I'd like to know a "break even" point for a typical carbureted vehicle. I suspect it's a fair bit longer than the ~7s cited in this video, and maybe more like a minute or two. I doubt anyone will do the research, though.
One last thing-a few years back in my city there was a big campaign about being "idle free." The time used in all the advertisements was 10 seconds. It certainly made me conscious of shutting my engine off in drive throughs and places like that. I even do it in traffic(on surface streets, not interstates) provided it's not too obnoxiously hot or cold. I don't personally own/regularly drive a vehicle with start/stop(although my parents' current vehicles have them) so any start/stop is manual. The MG only gets turned off if I think it's going to be a couple of minutes. Funny enough, though, drivers of city-owned vehicles seem don't seem to follow the "ten second rule."