I'm working a lot with sound (I'm a music producer) and I make videos with Blender.
One thing I have to do most workarounds with is the sound system.
There are several features that would greatly improve the workflow:
1. Add a VU meter. Make it a stereophonic master output meter with peak/RMS display (like the one Audacity has) - having a VU meter for every audio strip would be nice too;
2. Add a LADSPA effect rack that can be applied to the master output or any audio strip OR at least add a controllable brickwall limiter to the master bus (Fast Lookahead Limiter is a good plugin I use all the time);
3. Make crossfade effect strips work with audio too;
4. Get rid of opacity/blending settings in audio strips and put loudness/pan/effects in their place;
5. Enable changing parameters for all selected strips (display waveform, blending) not just the active one. Many times I need to make 20 strips display waveforms for syncinc, and I have to needlesly repeat my actions to achieve this while it could be done with two clicks (maybe a SHIFT key holded while pressing a checkbox could propagate the change to all selected objects?)
There are some things that I think would help us making videos with Blender

Do you agree with these? Or maybe these features are already done?