Time varying optimization wip

[cliffckerr]

@robynstuart could you pls do a partial review of this PR, which is running on sandbox, namely:

• https://trello.com/c/SPMrylzR/883-issue-1679-fix-daly-calcs
• https://trello.com/c/6eYMKULu/560-issue-1366-parset-should-contain-start-and-end
• https://trello.com/c/kwYAXRcy/950-fe-wishlist-define-cost-functions-page-programs-past-spending-and-coverage-data-consider-reordering-years-to-be-in-chronological
• https://trello.com/c/lsBgRdSK/900-issue-1696-outcome-mismatch
• https://trello.com/c/9mP1W8DB/946-export-data-for-budget-coverage-scenarios-fails
• https://trello.com/c/NetjgbmJ/947-cascade-plots-are-broken-for-scenarios
• https://trello.com/c/HUunNCxb/959-fe-sandbox-clicking-on-save-run-on-calibration-page-generated-radically-different-results-on-the-care-cascade-graph-from-what-ha

could you please not review:

• time-varying optimization, it's still broken.

other changes: general clean-up, including removing runmodel() (yay!).

finally, could you please update the hard-coded indices in plotting.py (around line 1000) with the values taken from project.settings? i started doing it but got confused so thought it'd be best if you did it.