I know that there are many people without university affiliation, who occasionally publish some papers,
sounds a bit like me (I'm still working, a lot actually - started freelancing/my own business which is still quite close to research. I'm chemist specialized at data analysis, so can do with an office. At the moment have an academic side job, though.)
Here's my experience:
There would be enough time to talk with collaborators,
As a lone wolf worker, you may be in more need for that than they are, though. Personally, after seeing former colleagues I'm quite happy that now I'm not subject to certain quite stressful experiences they have. Like having to write articles they are not actually, let's say, interested, in (or convinced about the scientific merits).
write articles I am actually interested in,
(somehow less than I thought I will - they still don't write themselves)
visit guest lectures at the local universities and the occasional conference on a holiday.
:-)
Apart from teaching,
(Well, I still do that as part of my business)
I would do what everyone else is doing, just on my own.
I'm far more independent than I was before. That is, I found that academic independence is mostly reserved to the professors over here but at the same time that any "career advance" above postdoc positions involve a change in profession from researcher to manager of researchers. For me personally, I thought it is easier and faster to achieve independence outside academia (starting a business is not without burocracy, administrative stuff or writing grant proposals [called offers] - but at least the financial results go into my own pocket).
But would it work the same?
(I think it's a bit different)
A possible doubt is, would many people even be interested in working with me?
I find they are. And if you are financially independent, probably even more so. I do have trouble to convince some academics that if they want my services, they should actually pay me.
I'm a reasonably competent mathematician but would probably be ineligible for many sources of funding,
Funding tends to think of employee researchers (or students) mostly - at least for the research part of a project. That's why I have that academic side job right now - a project grant where we figured that it would have been too much administrative hassle to try and get the existing funding released so they could hire me as consultat - so instead I'm employee with highly non-standard contract...
(Being SME industry is a funding category as well, though - but being an SME that provides research instead of getting research input is again outside the box)
couldn't offer a big name
(neither can I)
or any return visits,
I try to drop by colleagues whenever the occasion allows that.
and would be less motivated to get joint work published in the best possible journal.
I wouldn't say less motivated - it doesn't have the same prioritypressure as before.
For me also: if it is unpaid, then it has to wait until there's time for it. But most of my academic colleagues understand that - after all few of them would continue to work with the same priority if their institute didn't pay them.
However if a professor with permanent position comes along and thinks they can get significant amounts of free consulting and data analysis because they mention a co-authorship on some possible future "paper" I usually answer by sending an offer for a consulting contract which includes a paragraph stating that authorship for possible publications is decided strictly according to the DFG authorship rules (or COPE - doesn't matter, the rules are the same).
That's another interesting finding: some academics seem to think that the usual authorship rules don't apply. to me - presumably because I'm outside academia now (I had collaborations with them before - this is new).