Speed up checking for already downloaded files

[Aariq]

Currently get_prism_*() functions check for already downloaded files one at a time. This might be faster if all the files were checked at once. It looks like prism_webservice() acts on a single record and both retrieves the filename and checks if it is already downloaded. Maybe a solution would be to do all the HEAD requests at once to generate the filenames, then check which ones are missing, then download those one at a time.

[Aariq]

btw I've implemented something like this here: https://github.com/usa-npn/cales-thermal-calendars/blob/main/R/get_prism_tmean.R

[rabutler-usbr]

It's unclear to me how much this would speed up the download process. Given other workloads, I'm going to punt this to a different release.

@Aariq If you wanted to submit a PR with the change that would be great. Even better if it had some performance stats showing the speed up. Thanks for the idea!