In theory you should be able to do this with A Mic cue and an OSC cue with a long duration for each mic channel
The basic mechanism is this OSC query
/cue/{cue_number}/liveAverageLevel/{outputChannel} {low} {high}
Read-only; returns the RMS level of outputChannel. low and high are optional values to re-scale the output of this method. For example, sending the message /cue/1/liveAvrageLevel/1 0 100 will return a 0 if the level of output 1 is silent, and 100 if the level of output 1 is as loud as is possible.
You insert the query into the message that Resolume requires for whatever you want to control, using this method
http://figure53.com/docs/qlab/v4/scripting/osc-queries/.
Whether it will suit your purpose will depend on the refresh rate you get, and the latency. It would work fine for a clapometer type thing because that has a slow rise and fall time,and not too much rapid variation, whether it will be responsive enough to repond to complex, drumming patterns across multiple microphones is another matter.
Also remember that if you are using acoustic mics you will have some crosstalk.
I have a computer with QLab4 and Resolume on it, I'll knock out a quick test workspace tomorrow
Mic