Tracking issue for org TODOs

[dnephin]

This issue is to track all the TODOs we add while working on adding organizations, so that we don't forget to re-enable some checks once all the code is setup to support orgs.

• [x] ~https://github.com/infrahq/infra/pull/2896#discussion_r942780752 - require an org from the context~ - #2945
• [ ] - require all models to implement both SetOrganizationID, and IsOrganizationMember() bool.
• [x] ~https://github.com/infrahq/infra/pull/2902#discussion_r942792985 - require an orgID from the request~ This was fixed in a follow up .Replaced this item with another one below for the default behaviour when no org is specified by headers.
• [x] #2917
• [ ] remove the "set default org" from setupDB in server (and other) packages. This causes tests to not match production.
• [ ] add tests cases for all handlers to show that an a request fails if the requestor (access key or domain) is for the wrong org
• [x] ~https://github.com/infrahq/infra/pull/2902/files#r943705460 - setting of the org in data.initialize.~ #2921
• [x] https://github.com/infrahq/infra/pull/2902/files#r943713595 - test coverage for org_id migration - #2961
• [x] error if domain from host header does not match the orgID of the access key #2911
• [x] ~reconsider behaviour of unauthenticatedMiddleware when no host header or infra-org header are set. Consider removing fallback to default org. Make this configurable so that we have default org fallback for on-prem.~ #2945
• [x] replace the regex in SetDefaultDomain with strings.Map #2957
• [x] add tests for data.CreateOrganization
• [ ] return fake data for unathenticated GET routes when in multi-tenancy mode and the org is unknown