I answered a Nathan Bransford question, How Would You Handle the Query Deluge. My comment may have been a bit rude, but no disrespect was intended.
The first question to deal with when a literary agent says he is overwhelmed with queries from would be authors is to ask:
What percentage of the queries you currently accept actually make it to a published book?
If the agent's current query process does not generate a very high percentage of winners, then the current evaluation system in use is no better than a lottery. My suggestion to test this thesis is as follows:
Have an intern place slips of paper containing some query ID in a hat and draw out 10 each day/week. Follow up on the 10 and see if your percentage of winners increases, stays the same, or decreases.
If you determine that your current query evaluation process generates a string of publishing miracles, why change it?
