SciRate planning

[Jaiden Mispy]

Hi Aram,

I redid the landing page to a layout that seems generally closer to what we want, feel free to nitpick at it. (the specific data in top papers/comments will change)

We're approaching a state where I believe things are stable enough to consider taking over the live site and getting feedback from people as we roll out new features. The improvements I've made to the surface have been largely iterative and not transformative; it's still the same basic idea, but hopefully friendlier in lots of little ways.

The more substantial changes are in the backend. We now mirror metadata for every arXiv paper in history instead of just those that have been published since the site started, and have the architecture needed to efficiently perform complex search operations on that data. This ended up being a bit of a rabbit hole in terms of time-- that one search issue represents about 90% of the work I've done so far-- but I think it puts us in a good position technologically for later expansion.

There are some minor usability issues left that we discovered from random testing on people, and then I'm happy to deploy a beta and delve into the handful of older feature issues. We should probably discuss what to do after that; there's a bunch of potentially nifty ideas like https://github.com/draftable/scirate3/issues/133 and https://github.com/draftable/scirate3/issues/131