Popover or Sidebar Opt-in?

They do.

And in my experience most of those are people who have tested it for themselves.

Pop-up enthusiasts, on the other hand, in my experience, tend to be a mixture of (a) those promoting them, and (b) those using them who have noticed a short-term increase in opt-ins and made often-mistaken assumptions from that observation without going as far as split-testing for themselves in terms of their income over a reasonably significant time-period.

The point which many people miss is that there's an easy assumption that "a bigger list-size will necessarily earn more in the long run".

That assumption is often wrong.

As I found myself, to my surprise, when I split-tested it in each of four different niches, last year. In all four, I earned more (over a 6-month period) from the smaller list. At the time, I didn't quite understand why, but have now worked it out, and it's surprisingly simple: they're different people.

The easy mistake to make is to assume that "having a higher proportion of visitors opting in" can only be a good thing. If 20% more people opt in, the entirely mistaken assumption is that you still have all the "original" people, together with some additional people. This is typically wrong.

What you normally have, at least to some extent and sometimes to a great extent, is different people.

My experience, and the experience of others I know who have tested for themselves rather than repeating popular opinions which are often based on misunderstandings, is that buyers tend to be significantly among the people who'll opt in to a sidebar box, but be alienated by a pop-up and never return (which some of them would otherwise have done, so you need to monitor "returning visitors" carefully, too, and work out for yourself what proportion of them are lost by having a pop-up).

The people I know who have split-tested properly and carefully have abandoned their pop-ups. But that doesn't necessarily make that the right decision for you.

The one "constant factor", which anyone testing can reliably prove for themselves, is that there will always be people who, at the first sign of a pop-up, hit the "back" button and never return. Imagining that the increase in opt-ins is necessarily going to compensate for those people, in long-term financial terms, is simply an assumption. And pretty often a readily disprovable one by those willing to test, who think it through enough to appreciate that "numbers of people" and "long-term money earned" are not necessarily directly proportional, and that there are reasons for that.

Indirectly, this issue - like many of the Urban Myths of internet marketing - relates to the difference between quantitative and qualitative approaches to marketing.

Amateurs guess (and some propagate others' guesses as if they were factual); professionals test for themselves (and define the test-parameters adequately to take into account all the relevant variables). "Just saying" ...