Add support for ACS 2022 5Y

[jvtcl]

As with ACS 2023 5Y, the 2010 and 2020 PUMA geographies were recently reconciled for ACS 2022 5Y:

Source

We shouldn't need to manually estimate total group quarters in this case because that problem seemed to originate with the ACS 2023 5Y release.

[jvtcl]

The mixed geography problem does not appear to be fixed on the Census Microdata API:

• PUMAs are still denoted by PUMA10 and PUMA20 (Example query)
• PUMA10 and PUMA20 are still listed as identifiers in the API reference