How to get the best of the last half of a PhD?
Postdocs are typically useful for two types of things:
- They build experience as a (semi)independent researcher that is now typically required before you can be considered for a faculty post
- They are a chance to broaden or shift focus of your research
For example, I know somebody working in robotics whose thesis worked was entirely in simulation due to grad school circumstances, and used the "focus shift" opportunity of a postdoc to get into hardware. So a postdoc is worth considering if you feel that your current work is too narrow to be noticed by the type of organizations you might like to join.
As for where to go next: if you aren't absolutely driven to try to do a faculty job, don't try for it. There is a huge world of non-traditional scientific opportunities out there, which are often hard to be aware of from inside traditional academia. Some things to consider, with a lot of strange ecological niches that you can't even be aware of until you start talking to people in them. Do note, however, that if you aren't a US citizen or permanent resident, then many of these opportunities may be harder for you get due to visa issues.
For pursuing all of these questions, however, the big thing you need to start doing is networking. Get to know more professors in your department than your advisor, especially any ones who have had non-traditional career paths. Go to conferences and meet people. Look for events in your field that are also attended by non-academic researchers (this will likely be more in computer science or material science than physics), and look for the more informal and discussion-oriented attached events like Ph.D. symposia and workshops. If your advisor can help you get introductions, that's the best help you can possibly get, but since you're interested in a non-standard path they might not be able to.
As you meet people, be honest about your situation. There are always a lot of people trolling for good possible hires for non-faculty positions, both in academia and out of it, and there's a word of mouth network. If you've got good core computer science skills and don't come off as needy or desperate, there are a lot more possible matches out there than you might imagine.
With regard to:
I did research in a number of different field (...) finding out I'm not super talented, but I learn fast and I have some intellectual stamina.
I will try to focus on "talent" and "stamina", two self-evaluation criteria you mention which have not been directly addressed in the comments so far. You set up a contrast between stamina (which I interpret as a combination of curiosity and persistence) and talent (which I interpret as the ability to do more original/creative work with less apparent difficulty). You also mention several specific disciplinary sub-fields of research: xrays, etc.
I wonder if your perception of your "talent" might be more a consequence of the switching you have done between different areas of research, rather than your innate capacity for original research in one or more of these fields.
Research in learning and expertise development suggests that it takes about 10 years of concentrated work in a particular discipline (whether it is chess, car racing, or an academic field) to develop expertise. Once developed, such expertise may be mistaken for talent by a less-trained eye. Your words suggest a judgment formed by comparing yourself to others. If these others happened to be more experienced in a given area of research, then the difference in "time on task" may partly account for your perception of lesser capacity for original research in some ways.
A couple points seem worth making:
You probably do want to eventually identify, develop, and apply your talent in some field. Whether it is CS or Physics or something else, it has to be something you should be willing to apply yourself to for an extended period of time.
As you decide on your post-PhD path, consider the disciplinary area that you think could hold your intellectually engaged for a while. If your interest in research is such that CS is a means to answering questions in other fields, then CS is a means rather than an end for you. This will take you down one particular career trajectory. However, if your interest is in questions in CS proper, this might imply a different trajectory (e.g. a research scientist/programmer in a physics lab, vs. a research scientist in a research center that focuses on computational questions like algorithms, etc (sorry, CS is not my area).
Finally, saying that you "like doing research" is a starting point for some deeper reflection. Think about HOW you like to do "research" - stated differently, what are the characteristics of a research project that make it exciting and fulfilling to you. A big part of this has to do with the level of "agency" you are comfortable with. Some people find satisfaction in carefully nurturing their own research agendas. They do this by working in a very specific niche for a very long time, spearheading (as a PI) their own grants and building partnerships and communities around a specific theory and/or methodology. Others are less comfortable leading, and prefer following -- rather than agonize over what grant to apply for, they are happy supporting on-going work, where the decisions about the direction of inquiry have mostly been made for them. While both kinds of people may (justifiably) consider themselves "researchers," the differences in approach can lead to very different career trajectories in terms of the role one is shooting for and the kind of institution (size, mission, etc.) that is compatible with each of these roles/goals.
Good luck!
From what I can tell from your question, the industry seems a logical choice. For the remaining time, I'd advise you to look into branches that might interest you and then focus your development on skills required there. There is a plethora of various opportunities for your type of background, whether in the research sector or in engineering.
That being said, if you don't plan to stay in academia, doing a post-doc seems not to be the best investment. People usually do a post-doc in order to improve their chances for a professorship (or as a kind of placeholder until they get one).