The table shows that weeks, days, hours, and minutes are equivalent to the UCUM units, while seconds and milliseconds are equal to the UCUM units.

Based on my reading of the CQL and UCUM specs, at least hours and minutes should be equal. See UCUM Section 31 which defines:

• wk as 7 d
• d as 24 h
• h as 60 min
• min as 60 s

In CQL Appendix H, it says "in CQL, an hour is defined as 60 minutes" and "in CQL, a minute is defined as 60 seconds" – so since CQL and UCUM seconds are equal, and since they both define hours and minutes in relationship to each other and seconds, those all should be equal.

CQL day and week are tricky since their actual length is context-dependent (e.g., a day when the clock changes may have more or less than 24 hours) – so perhaps this is why the relationship is defined as equivalent for these? I'm down with that, but if the quantity is devoid of context (e.g., define OneDay: 1 day) then am I right to assume that definition equals `1 'd'`?