Validation: develop a strategy for using different data models

[emi420]

Currently, Underpass can identify features with tags that are not in the data models defined in YAML configuration files

This is useful for running an instance of Underpass configured for the specific user needs.

But there are cases when a more dynamic reporting is needed. For example, at HOT we have a lot of mapping projects, some of them remote, others on the field, and for a range of impact areas. If there's a mapping project with a specific list of tags that must to be followed, we need Underpass to be able to show which features are tagged using tags and values that are not on that data model, but without changing the YAML configuration files for that.