this certainly makes sense as an feature, it's in the standard, and it's reasonably easy to do. But I'm not sure it's a subtask of application-period tables, looks more like an independent feature to me.

to see also 4.14 Periods, 8.20 <period predicate>

This comment could come too late to be even considered, but...
Wouldn't it be better to add range types, indexes and generic syntax to handle them, like PostgreSQL? (not necessarily compatible syntax, as pg syntax is a bit questionable)

https://www.postgresql.org/docs/12/rangetypes.html

yes, it would be better. Periods are just coming together with SQL standard temporal features — system-versioned and application time period tables. On itself, standalone, they aren't a particularly useful feature.