Emotive is not that more complete and advanced : it is just more opaque, and marketed.
I don't like the kind of policy of theirs, although it kinda works. But I wouldn't recommend it.
If you want to use such device, I'd recommend openbci way more. It requires a bit more that you put your hands into it at first, but...
it's open !!! So it is likely to evolve thanks to the community, especially the hacking one (emotive is more for marketing shit).
For an artist, it should be a lot more rewarding in the future.
Usine is fantastic for this, although some pre-processing is necessary, for which Usine might not be that handy (in the context of OpenBCI, Processing is used for hte interface, and you can tweak from there, so that you can send pre-processed data to usine, such as energy per frequency band for example).
Now, don't fool yourself : brain signals are very noisy, if not messy. It is difficult to isolate "markers" that make sense both to you and as a quantified signal. Thinking deep about quantum physics or about a donut won't bring much difference really, but if you cough, your signal will explode.
The idea is attracting, but the reality is difficult to handle. But that should evolve in the coming future.
As for people using it in music projects, there are some definitely.
Although I started to experiment with it some time ago with a friend of mine, I can't find the time currently (I do own an openbci). I should pursue this soon I hope.
The pioneer in that domain is Alvin Lussier of course, who did so half a century ago (
https://www.youtube.com/watch?v=bIPU2ynqy2Y).
I've seen many recent projects using it, and currently the project I'm the most interested in seeing development is eegsynth and their collective "1+1=3", very nice and brilliant people I met at a brain hacking event :
http://www.ouunpo.com/eegsynth/?page_id=530
http://www.oneplusoneisthree.org/#workshops
(haha, didn't realize my EEG headset is somewhere on those pages).