Their point is that there's absolutely no effect of banning travel between two areas with similar prevalence, based on rather simple logic.
Take two bags with mostly white and a similar number of red balls in them. Then, twenty times, grab one hand from each bag and empty it in the other. What's the expected change in number of red balls?
Multiple cities with hundreds of thousands of people in both the US and Europe are infection free right now. Extending that for even a few weeks is very useful.
By your model, some reductions in travel ought to have a negative effect. Is that a fair corollary?
Limiting all Schengen traffic is an extreme example, but it's not obvious that some level of restriction would necessarily produce a significant bottleneck effect (given that citizens are exempt).
By your last analogy, the red balls would have paint rubbed off them in the exchanges.
When the concentrations in the US get higher it will start to make sense to constrain travel more over there. Right now the chances a domestic flight contains an infected individual is quite low, but that won’t always be the case.
Bear in mind the virus cannot be stopped, or contained at this point. All we can do is slow it down while also limiting the harm our countermeasures cause.
Saying the virus cannot be stopped seems defeatist. China seems to have managed it locally, albeit with some very aggressive measures.
That's what you're proposing, if you don't realize it. If you get the growth rate below 1, you can wipe out the disease, instead of merely delaying it.
But this is clearly false because population density is another important factor for contact between people.