Feature request: workout deletion/simple edits via API

[dwildstr]

This is a suggested simpler stopgap (and also enhanced version for power users) feature for batch work as proposed in #514, which understandably would require a fair amount of UI overhaul. If the API had calls suitable for deletion and editing of workout metadata (i.e. text fields like "type", not actual point-level GPS data), then client code could be written to perform batch operations client-side with multiple edit/delete operations.

For instance, I'm wanting to delete every workout under 60 seconds operating under the assumption that it's a glitchy recording. With batch operations in the UI and sorting by duration, this would be pretty easy, but it would also be easy with a Python script using the workouts API endpoint and a notional delete endpoint. Likewise, if I wanted to ensure that, say, any workout between noon and 15:00 had type running, regardless of what type of exercise it's guessed to be, I could write a similar script using the notional edit endpoint instead.