Both arrays are sized, as in this post, to make the same energy over a year: a 1 MW tracker against a larger fixed east-west array. Pick a city and a day to see their output, then set the module price to see which costs less to build.
Prototype. Array sizes and the non-module costs come from the post's tables (east-west $0.31/W, tracker $0.52/W all-in apart from modules, so California reproduces the post's $633,608 vs $681,218). Daily curves use a simple clear-sky model with each city's cloudiness tuned to the post's sizing ratios; for Buffalo the model can't get below a 1.09 ratio where the post finds 1.00, since snow and diffuse light aren't modelled. Expected output on an average-weather day, not a real weather year.