"Project" level views that are neither project_name or subject_id alone

In the effort to replace spreadsheets with proper metadata-based forms I had a discussion w/ Kenta and Rachel about this spreadsheet used to track mice doing fiber photometry, loosely defined as "Kenta's mice"
https://docs.google.com/spreadsheets/d/1RhL4-34iuBMbMLKc__6DN_LqJ0Xd9DP_r4z7rZn-rgk/edit?gid=1981155024#gid=1981155024

There is no single field in the metadata that can aggregate all of these mice cleanly since they span projects and even include mice that have missing metadata, etc. But we can still come close to replacing this spreadsheet by writing an acorn in zombie-squirrel that grabs this metadata. We can create the tab separation by adding a grouping column that splits by project name.

Query is a mixture of project names and filters for subject_id. Assets should be aggregated by raw > derived groupings.

- ID: subject_id
- DForge: dynamic foraging start and end dates (how does one know the end date? after mouse died?). Can be taken from the acquisition_type or by going into the stimulus_epochs?
- Pavlovian: same as above
- VRforaging: same as above
- CS2: same as above (task variant, request users put this in the acquisition_type?)
- DOB: subject metadata
- Sex: subject metadata
- Genotype: subject metadata
- Status: tricky -- we have to combine genotype with injection information to compute this. This also requires the intended_measurement information.
- Surgery date: procedures metadata
- NSB: presumably the NSB experimenter name?
- (more todo)