I hate RFPs with a passion. If you take them at face value -- and I'm talking about RFPs for software systems or development -- and give them everything they ask for they invariable end up buying something that breaks all their own rules.
I'd never limit things to just replying to the questions they ask in the RFP. I see it as the first contact, and an opportunity for me to give them an idea of what I can actually do for them.
It should definitely be followed up with emails, discussions, possibly phone calls. I don't think I even give enough information in the proposal for them to make a final decision, unless they have no idea what they're doing.
I have also had the opposite experience of spending a great deal of time convincing them that they are asking for the wrong thing only to lose the job to another company that is perceived to bigger or more experienced at doing the very thing I have recommended.
Or after all is said and done you realize that the winners was picked ahead of time and the RFP was just a process to satisfy some bureaucratic requirement.
It is very difficult to tell the difference between a well answered RFP and a paid consulting report I have never done this, but every time I allow myself to be sucked into one of these situations I swear that next time I will demand a consulting fee for participating.
In fact I was involved in a very large government related RFP fairlyrecently. At the end of stage one there were still more than 50 vendors on the "short list". I refused to participate in phase two with out a consulting fee or compensation. I didn't get it of course but I probably saved myself a great deal of time and effort.