use PS64 to resolve Rydbergs, add denser model
These adjust the number of l-resolved shells according to the figure in PS. Also add denser model, log n =8. The ratios in the spreadsheet are today/Porter and they are much much better at lower densities. There are major differences in 10830 due to the LS CS bug Giulio found (those cells have a red tint) and the strong diagnostic lines have a yellow tint. Those are good at low densities and diverge at higher density, probably due to the different n-changing collision theory we now use.