Model table update

[pgm]

In 24Q4 some columns changed names and some additional columns have been added. Since the model table consistently changes from release to release, I changed strategy in this PR. The front end will always have a dependency on the schema of model.csv. However, that doesn't mean all of the portal-backend code needs to as well. Instead of mapping column names from model.csv -> snake case -> portal db -> json response -> react component, I've cut out several layers.

Now, any fields in that the portal depends on, should still go into the database, but for all columns, with column names as they appear in the Model.csv go into a single field as a json encoded dict, which the frontend can receive. There is also a interface automatically defined to ensure the frontend is kept in sync with the columns in the actual model.csv file.

This means that for columns that the backend doesn't care about the flow is:
model.csv -> db -> react component

Which eliminates a lot of boilerplate.

(Also this PR includes some small fixes to react warnings I saw when testing)

[snwessel]

Looks good to me (and much simpler!)

I appreciate all the explanatory comments.