Hello Ed,
Fret not about the 16" Meade being out of your budget; I wouldn't recommend it anyway.
I'm not going to give you much specific advice here, but here is what I've learned putting together a couple dozen robotic facilities:
1) Spending lots of money on a premium mounting is generally a waste if you are going to carry a telescope smaller than 16" or so. The reason is that the software is capable enough to bring, for example, a cheap Meade 12" LX-200 mounting into the same or nearly the same performance level as, say, one of the AP GTO mounts. By way of real-world example, the 12" LX-200 at 933 has a median actual pointing error over 100 slews totaling 1500 degrees of 9.1 arcseconds; the AP mounting at a site I've helped out at near Cloudcroft did 9.0 in an identical pointing test. We used USNO A2 to solve for pointing, so the difference in this result is well below the measurement noise. I've got a couple dozen examples of this kind of result, and you are welcome to ask around to the people I've worked with about this; so the lesson for me is, buy a cheap mount, and buy good software, if you are on a constrained budget. (Note also that for about the price of one of the premium mounts you can get two 12" Meades and use one as a hot-swappable backup; this is an option that a few people have taken.) A similar argument holds for tracking; I hear people with premium mountings talk about how they can track for eight minutes without using a guidestar, and that is sometimes true, and when it works it is very impressive. But what they don't tell you is that every time there is the slightest breeze, the FWHM of the stars in the image doubles. This is the 21st century, not the 1950's: software can easily and quickly grab a guide star for you, even near the galactic poles - its dumb not to do it. It solves a lot of problems that even premium mountings have.
2) On the other hand, if you are going to mount a > 16" scope, you need either to make it extremely light (and then you do need the load-carrying capacity of a premium mount) or else you need to mount it on something made for the telescope. When you get into this range, you are pretty much looking at PC-TCS, DFM, or similar (generally ACL-based) control systems. Either way, you are putting a big cross-section to the wind and the number of observable nights will go down. Or else you are going to a skeletal tube structure and the background values in your images will go up. (I'm not saying this is necessarily a bad compromise, but a lot of people don't think of it.)
3) I've seen a lot of high-end mountings that don't slew accurately, and I've seen a lot of cheap mountings that don't slew accurately. In most cases, I've found that carefully balancing the mount improves things substantially, whether or not you are running an error model. If you end up with a fork, make sure you balance all axes (l-r fork, front-back fork, front-back and top-bottom tube, at least) and maximize your moment. People who are slipshod and don't do this often conclude that the mountings they use are bad, but what it really shows is they don't know much about engineering a mount. Now if someone wants a high-end mounting, for whatever reason, that's fine - I'm not anti-fine-mountings here, I appreciate them very much. But if someone gets one for $4000 extra, to accomplish with brute force what could be done for a couple hundred dollars of weights, well, that's dumb and amateurish (in the worst sense of the word) in my book.
4) Speaking of error models, they are great. You should definitely use whatever terms are needed to null whatever pointing problems you have. But, you should *only* use an error model *after* you have optimized every mechanical system to result in best-possible pointing without the model. Any other use of a model is a crutch, and eventually it won't leave you satisfied.
5) The telescope that would best fill your needs is probably *not* a 10" to 12" Ritchey. I say this not clearly knowing your needs, but, in general, most peoples' needs in the smaller sizes are best filled by a Newtonian. There are several reasons. One is that they are easy to have made in a given focal length to get just the plate scale you would like; this obviates the need for a $400 SCT reducer-corrector or a $1000 R-C reducer-corrector. Another is that correctors for newts (e.g., coma correctors) are plentiful and dirt cheap (unless you are using a 3" chip). Also the telescopes are much less expensive per aperture than a Ritchey. SCT's also have a rather significant cost advantage, with the disadvantages that go with that (some quality uncertainty, typically); they are much cheaper than a Ritchey up to and including 16", especially if you consider the secondary optics issues, and you can't argue with the results people have gotten from them. Where a Ritchey with a flat-field corrector will really help is if you spend $60,000 on a CCD camera with a really large chip, OR if you are going to shoot film larger than 35mm. Both 701's 20" and 32" R-C can fill a 7cm frame with light, and medium format film was part of the design spec there. But if it hadn't been, buying a system with a 3" fully illuminated field, expensive secondary optics, and gigantic secondary obstruction, would have been rather wasteful.
6) Whatever hardware you end up with, you shouldn't consider anything that anyone else says to be gospel. Your analysis of your budget and your usage scenarios should drive the hardware you purchase. I do recommend visiting some successful sites that use hardware similar to what you've chosen before you place your order. This can be very educational and useful.
7) The exception to the above is this gospel: Get good software. Don't skimp. Look at what the successful stations are using, and get that. Be a pioneer if you want, but that doesn't maximize your chances of success.
>Also, I am wondering what is a really good CCD to match
>the capabilities of a scope in this price range.
There is a brief article about getting the proper image scale (for research purposes) at:
http://www.roboticobservatory.com/information/image_scale.htm
>Given $10K, what would be the best, most
>versatile setup? Given $5K (which is much
>preferred), what would be the best, most
>versatile setup?
Shooting in the dark (you give almost no usage scenarios), here's what I'd do at those price points:
10K: Build a newt, build a mount, automate it with a Bartels drive system, get a cheap SBIG camera.
5k: Build a newt, build a mount, automate it with a Bartels drive system, build a CCD camera.
Both of these things have actually been done at these budget points, and you said something about electrical engineering being in your background. So don't think you have to double or triple your budget just to get something that works - you don't.