grcq should become gamlss fits in shiny apps

[mcfrank]

We found that gamlss gives much better fits and outputs percentile ranks really nicely. it's a little slower but a lot better. here's the demo that could probably be adapted pretty easily.

https://github.com/mcfrank/demo-vocab/blob/master/gamlss_demo.R