I'm not Tim, but... <g>
Depending upon the circuit that drives the VU meter, the rectifier inside the meter can cause odd loading of the audio circuit and introduce distortion. Hence the need for a buffer amp to eliminate that chance.
Here's a circuit I culled from previous projects that will nicely iso the audio path from the meter.
http://www.brianroth.com/library/vu-buffer.pdf
It "looks like" a balanced center-tapped 20K line input transformer to the driving audio circuit, and uses cheap parts.
As for the meter "ballistics", there's an old ANSI spec that describes how much time the meter can rise and fall to a test-burst signal, how much needle overshoot is allowed, etc.
I think Jay McKnight has the test procedures at the MRL website. If not, I have a copy of Jay's test methods here somewhere.
Bri