Investigate whether there's a benefit to caching works/editions in bulk

[cdrini]

Although caching internet archive API results in bulk is definitely faster, there might not be a benefit to caching the work/edition objects themselves since they'll only be asked for once.

[tfmorris]

The database also maintains a cache which it can use for repeated requests.

[tfmorris]

Didn't you already implement this? It looks like IA requests are batched (which is definitely a win since you only pay the R/T latency once rather than N hundred times).

[cdrini]

This is caching hits to the postgres copy of the database, not to IA. The question is whether to remove the postgres caching.