Nephron data processing

Have no idea what this is called in real computer science but giving it this name for now.

An app faced with the problem of caching data that is retrieved from a source has 2 choices.

(1) They can create a model that reflects the data to be stored. A dynamic approach.
(2) Or they can pick the data they want according to a static model and ignore the rest (to urine exactly like the nephrons of the kidney).

(1) has the advantage that the designer doesn't need to actually code the model. A list of data in the cache can be produced generically so that at no point does the coder ever need to know the details of the model. This is wonderful if the model is likely to change in the future. However it also means that all garbage also gets processed by the system, and introducing malicious contents is easy (once encryption security on the data channel has been compromised).

(2) A static model makes more work for the coder, and it makes the app more rigid and inflexible. Changes to the model will not be shown up in the app and will require a new version. But it has the great advantage that the end user is better protected from garbage and malicious intent. Any data the consumer doesn't expect is sent to urine, like a toxin in the blood stream.

I just opted for nephron data processing myself hence the post.