A pretty common complaint among argument theorists about the fallacy of begging the question and circular argument is that hardly anyone ever really commits the error. But then there are the cases where it happens for realz.
President Trump, before flying to the G7 conference in Montreal, argued that Russia should be included in the proceedings again — so, returning the meeting to the familiar title, G8. Reported at Politico and InfoWars(don’t read the comments!) (Russia was expelled after their 2014 invasion and annexation of Crimea.) Here’s the argument:
I would recommend — and it’s up to them, but Russia should be in the meeting, it should be a part of it. You know, whether you like it or not, and it may not be politically correct, but we have a world to run . . . And in the G-7, which used to be the G-8, they threw Russia out. They should let Russia come back in. Because we should have Russia at the negotiating table.
As far as I can see, the explicit form of that argument, given the ‘because’ clause, is:
We should have Russia at the bargaining table
So: Russia should be part of the G7(8).
That’s pretty much a perfectly circular argument, since the premise is just a differently worded version of the conclusion. I think the only mitigating factor to this fallacy challenge is that Trump also says, “we have a world to run,” which I think is a point about economic and political necessity. Something like: Look, Russia has been and should be sanctioned, but leaving them and their economy out of these discussions is short-sighted…” But he doesn’t do that.
One lesson, then, is that fallacy charges of circularity may be good ways to elicit submerged reasons.  Like what we see with the Trump case here — there is a hint of a better argument in the background, but it’s really just a series of assertions of the conclusion. The charge of begging the question is a way of getting those other reasons out for evaluation. So there’s something right about the argument theorists’ complaint that there aren’t really circular arguments, but there’s also something to the thought that the fallacy categories are useful.