Equate exposure and visit dimensions

PFS has historically used visit to refer to the exposure index, and the quantity of support code relying on this demands we support using visit rather than (or, at least, as well as) exposure as the exposure index.

Jim Bosch suggests the following two potential plans of attack:
Plan A: Keep exposure and visit as defined by Rubin. exposure would still appear in some data products, but (like Rubin) it'd be just be calibrations and ISR, so science users would only ever see visit.
Plan B: Drop exposure entirely and only use visit, modify RawIngestTask to do s/exposure/visit/ (I'm sure we'd accept a PR that made that configurable; it may already be a module constant). This seems cleaner in the end, as long as the s/exposure/visit/ work on things you want to reuse from Rubin starts and ends with RawIngestTask.

Time pressure (the next run starts tomorrow night) has me thinking to try Plan A.