I'd like to use WaveGain to adjust volume on a wav before encoding to mp4. Should I dither or noise shape the files?

If I read other posts correctly, I shouldn't use dithering or noise shaping if I'm going to process the file after using WaveGain. Is this right? If so, is encoding considered "processing" to that I shouldn't use dithering or noise shaping?

It would be easier if there was a replaygain for mp4's.
I'm using quicktime pro to encode, if that matters.

Yes, you would be best advised to dither the output, assuming you're outputting at 16 bit. In this instance, WaveGain is considered the last process and if you don't dither, you will lose resolution. If you were going to take the output and further process it through a sound editor, for example, then you should output without dither at a higher bitwidth - 24 or 32 bits. Personally, I would also recommend using noise-shaping, but it wouldn't hurt for you to test this yourself.

Thanks

What level of noise-shaping would you recommend?

Probably light, but I do suggest you test them out with whatever you're encoding.

any of the AAC encoders out there support a --scale setting?

None that I know of.