Resolve issues with nuclides management

[youngmit]

A handful of tasks already exist which raise issues with the nuclide bases system, which can all be addressed somewhat in tandem:

• Global state is bad: #473
• Nuclide settings shouldn't be in Blueprints. We shouldn't lose sight of the fact that this is essentially the only reason why blueprints are needed throughout ARMI for anything other than building a Reactor model from input, or whipping up fresh assemblies when shuffling fuel. Pulling the nuclide directory out of blueprints and onto the reactor itself will be huge.
• It would be nice to split the nuclide directory stuff out of ARMI entirely: #259 and #460. To accomplish this, we need to make them re-usable, which largely means tackling the above.

[john-science]

As a matter of "good design", I would also like to point out this code:

https://github.com/terrapower/armi/blob/fd5d2dd31051206ce62696027f05a9304b499677/armi/reactor/blueprints/isotopicOptions.py#L514-L538

It seems a little inside-out that we have to make definitions here that relate to specific, external programs developed by other teams.