I'm no expert, but the main decision for anyone who wants to image is to choose the best CCD chip to suit their needs. Note that this decision is *independent* of any of the camera manufacturers or the "body" that the chip will go into.
The chip choice will depend on what imaging you want to do and even though there has been a profusion of new (and larger) chips in recent years, I'm pretty sure that most people will need to compromise on at least one of their purchasing decisions.
I terms of the chip then, the key parameters to decide on would be pixel size, ABG or NABG, mono/colour, QE and the propensity for the chip to "suffer" from RBI (residual bulk image).
Of course, the weight you put on these parameters will depend on your imaging requirements.
For pleasure imaging, you'll probably want an ABG (anti-blooming gate) colour CCD that gives you a well sampled image (related to pixel size and your imaging train optical characteristics). But even within that, if you're imaging planets, you can push the image scale a lot. For deep sky imaging, you don't want to oversample too much.
For astrometry, image scale is really important so matching the pixel size to your optics (and typical sky conditions in FWHM units)is important. However, you can throw in a focal reducer to help out if necessary.
If you want to do photometry, then you will certainly need a NABG chip and you must limit the amount of "glass" in the optical train - focal reducers, non-essential filters etc. RBI is really important (or rather, lack of it) and you want as high a QE as possible. Microlensed and colour CCD's are pretty much out.
As I say, I'm no expert and there are a good few people out there who know more about this than I do. All I can recommend is you decide what you want to achieve with your setup (pleasure or science) and then make your CCD decisions based on that.
One final point, even if you choose pleasure imaging, you can still do some science. Its not completely ruled out but choices on what science you *can* do will be limited. For example, photometry of anything other than large amplitude objects is out. You won't be constructing a curve for an exoplanet... But astrometry and some photomeetry projects would be do-able.
Equally, if you concentrate on science, you can still take pleasure images. And typically good ones at that.
It seems to me that using a pleasure imaging setup for science is limited, whereas the journey from science to pleasure imaging is much easier. The fly in the ointment is cost. A science setup would usually be significantly more expensive.
A bit long winded, but hope it helps.
More on image scale:
www.stanmooreastro.com/pixel_size.htm
  