"They ramble, someone has to move it on"

Here's the objection from the practical end. Some people go on. If you don't cut in, they'll take the meeting into next week, and someone has to be the one who moves things along.

Two things. First, the timing evidence again. In the medical studies, the fear that patients would go on for ever turned out to be wrong by a wide margin. Six seconds to say what they came for, when nobody stopped them. Marvel's patients who were allowed to finish used six seconds more than the ones who were cut off. The doctors were interrupting people who were about to stop anyway. I'd expect the same at most tables and in most meetings. The person you think is rambling is usually a sentence or two from the finish, and the jump arrives just before they'd have got there.

Second, and this is my reasoning and not a finding, jumping in trains the thing you're complaining about. A person who expects to be cut off learns to talk faster, to front-load, to pad the start so that the point is protected, and to keep going through the gaps because a gap is where they get taken. If you've been jumping on someone for years, some of their rambling is a defence you built. Let them finish for a month and see if it changes. In my experience it does, and quickly, because most people would rather say the point once at a normal speed than say it three times fast.

There are real monologuists, and there's a proper way to interrupt one, which is at a finish, by name, with a question. That's chapter 23. But be honest about how many of the people you jump on are that, and how many were just about to stop.