The only "feedback" you ever get on a query is in the form of requests for more material. The more queries you send out, the more requests for partials etc. The drawback with sending out 1-5 queries is that it takes you a lot longer to figure out if your letter is working.
I started by sending out 20 letters to an initial mix of dream and meh agents. I got a very good response (30%) right off the bat, and so I continued to send out an average of twenty a week (email, when possible) and ended up querying over 300 agents over the course of the 8 months it took me to finally sign with an agent.
I would keep track of who was selling what on Publisher's Marketplace and which agent was moving where or going solo on MediaBistro and AgentQuery to keep my list up to date.
Of course, I thoroughly proofed everything I sent out, email or snail. And I did personalize every query, adjusting my approach based on specific info I might have on any particular agent. I never queried more than one agent at any agency. In order to keep it all straight, I set up a spiffy excel spreadsheet to track progress.
It is a time-consuming effort - a lot of hard work and $ for supplies and paper, but I decided that in my case (unpublished, uncredentialed with absolutely no contacts in the industry) I needed to improve my odds by querying as many agents as I could. In the end, I got three offers, and the agent I went with got me a two-book deal, so this method worked for me.