My personal experience is that for best results in the field of CCD imaging one should load the mount with only 1/3 of the nominal payload when imaging without or limited wind protection and a mobile setup. If you don't suffer from wind shake and if your mount is permanently installed on a pier, you can go up to half or even the full load capacity. How much that is depends largely on the quality of your brand, i.e. how realistic the estimated max payload is.
If you are serious about CCD imaging and want best results, don't save money when buying the mount. As a rough guideline, spend 1/3 of your budget each on the scope, the camera and the mount. If in doubt (you better be), spend more money on the mount and get the next larger model.
Normal polar finder scopes usually work well enough with CCD imaging, since exposure times are short. Drift alignment is imho not necessary.
With a normal polar finder telescope my field rotation rate is on the order of 0.008 degrees per hour. That corresponds to a shift of 0.3 pixels per hour in those corners of the ST-10XME that are furthest away from the guide star. Hence, for my setup even for 30 minute exposures drift alignment is not necessary.
With an autoguider (AG), the periodic error is not a matter of concern as long as it is smooth and slow so that the AG can correct for it. Then the amplitude can be 10" or even 15". It also depends on the AG's integration time, on the focal length you are imaging, and on the maximum guiding error you are willing to accept. Periodic error correction (PEC) is superfluous if you use an AG. Switch it off, unless your AG exposure times are very long.
If your tracking error is erratic and with fast outliers, and you can't fix it, then you could decouple the mount from the camera with a tip-tilt device such as SBIG's AO units. If this is not an option, think about reducing the focal length such that the periodic error is not visible in the exposures.
Always choose a star in the middle of the field of view for focusing. If the camera is slightly titled against the focal plane, or if optical abberrations are signficant, you will get a very uneven image quality across the field if you do not focus in the image centre (the optical axis). The integration time for the individual focus exposures should be a few seconds so that the seeing effects can average out.
Do not rely on the same focus value if homofocal filters are used, unless you verified it for your setup. With fast optical systems the focus tolerance can be as small as 0.01 millimeters. That translates to a tolerance of about 0.03 millimeters for the filter thickness. Optical abberrations can also lead to wavelength dependent focus values.
The better the seeing is, the more care should be taken during focusing. This holds for exposures in all filters.