I've also added a benchmarking utility so that I can have a fancy README with all the run times 😎

It's always DNS:

> We have identified a potential root cause for error rates for the DynamoDB APIs in the US-EAST-1 Region. Based on our investigation, the issue appears to be related to DNS resolution of the DynamoDB API endpoint in US-EAST-1

Ah, right. Yes, I see it could make a bit of a difference:

>>> from pympler.asizeof import asizeof
>>> asizeof(Case(When(pk=1, then=1)))
3408

Wouldn't raw SQL strings take even more memory?

>>> sys.getsizeof(Case(
... When(pk=1, then=1),
... output_field=IntegerField()
... ))
48
>>> sys.getsizeof("""CASE WHEN "your_table"."pk" = 1 THEN 1 ELSE 0 END""")
90

Yes, it's milliseconds. The web server and PostgreSQL server run on the same machine, so there is no network lag.