I too am unsure of what part of the process is the most troublesome. Assuming it is the extraction my opinion FWIW. Getting it right and as close as possible looking good in camera by having an image well defined and seperated from the background is half the battle before worrying about particular tools and how to extract the image - there are many methods.
IMO Green or Blue screen can be more trouble than it is worth for stills, great for TV and video though. The main issues seem to arise from not lighting the background adequately, not enough seperation between subject and background leading to troublesome colour spill, needing yet another correction.
A mid grey screen may be all you need properly lit seperately from your main subject. Depending on the subject it is sometimes possible to dispense with hand building a mask and instead use layers, blending modes to isolate your subject then remove the background by various methods.
For small objects in a studio for example (fully controllable lighting) you may be able to get enough seperation between subject and background to take two images one of the subject lit to taste and one of the subject as a silhouette. The silhouette serving as an accurate mask when the two combined.
Depending on the complexity of the subject outline the tools of choice may be either the Pen tool or the Extract tool (I think both available in CS2)