Factor out/minimize overrides files.

[samreid]

Discovered during #299. The overrides files must be maintained along with the simulation, and things will be simpler if we can minimize them as much as possible.

This can be done safely by comparing grunt generate-phet-io-api before and after a change. I'll take a look.

[samreid]

I've committed several changes factoring out overrides from several places into the point of origination, I was able to do around one per minute (I focused on the ones that did not require simulation-specific knowledge), and it seemed a good use of time. For instance, atomic interactions is down from 27 to 19 overrides. I recommend taking a bit more time to look through and see if anything else is easy to factor out.

[jbphet]

@samreid - I have a few questions, and a comment, on this:

• It looks like you moved some overrides and not others. In your comment you said, "I recommend taking a bit more time to look through and see if anything else is easy to factor out." Is the goal to move them all? If so, what is the purpose of the overrides file? If not, what are the guiding principles for what should be in the overrides file and what should be in the code?
• I ran grunt generate-phet-io-api as you suggested about, and it generated a file that includes a number of differences that look like they come from common code. Also, the format of the file name seems different from what's in the directory now, and the file seems to include a date a time in the name (2020-4-30_11.20.41.json versus 6-30-2020_1.39.04_PM.json). Am I supposed to check these in on changes, or is the idea just to compare them and look for the changes that I intended to make (and to ignore things from common code)?
• Finally, just a heads up: The changes for this caused CT errors in atomic-interactions and states-of-matter-basics, so if you do anything else on this, please make sure to check those sims as well. See CT Any schema entries in the overrides file must be different from its baseline counterpart atomic-interactions#87 and CT Any schema entries in the overrides file must be different from its baseline counterpart states-of-matter-basics#28.

[samreid]

I reverted the changes and used grunt generate-phet-io-api to confirm that the overall API was not disturbed. @jbphet can you please review?

[jbphet]

I reviewed this by skimming through the overrides file history for SOM, SOMB, and AI, and they look pretty reasonable to me. The only thing in them is the phetioFeatured settings and they look pretty close to what I remember. The featured items will likely get reviewed before final publication too, so I think this is safe to close.