This pull request fixes a few bugs in the steam Rankine power cycle models:

• Improves calculation of normalized cycle performance from nondimensional tables
• Removes artificial lower limit on air-cooled condenser pressure
• Insures that cycle model and air cooler model assume same design point approach temperatures and calls performance model during cycle design method to verify it generates the design point performance and calculate design point fan and pumping power.

The net result of these changes increases the annual energy of the default molten salt power tower technology model in SAM by about 5%. Other technology defaults have a smaller response - see the test results file in the SAM repo.

This pull request also begins to transition the molten salt power tower user interface to define calculated variables through a callback to the power tower compute module. While a few input pages still use the equations window, all MSPT cmod inputs are either direct user inputs from the UI or translated from widget values (e.g. HTF code translated from 0-based array to fluid code corresponding to drop-down labels)