I'm trying to solve for a scenario where we want to leave a document containing (destination) links exactly as it was at the time it was finished and published (off-platform), to ensure traceability, allow for re-use, and to avoid confusion. E.g., let's say we did a certain report in 2022, linking back to our factbook. The factbook values (at the source) are still expected to change over time, but we want this report to basically be "frozen" as it was without seeing new data (2023, etc.) appear in the report. (Obviously we could basically stop updating that factbook and start using a new factbook with new documents, similar to one standard roll-forward model, but that's undesirable for a few reasons and we don't want to be forced into that model.)

The problem is, there is no way to "freeze" link content. Unlike connections, since the power of updating link data is at the source ("pushing" changes, i.e., publish), the downstream destination is unable to prevent that updated data from coming in. This is especially concerning since, by design, in a factbook the dates may be updated, meaning that a column holding values (and links) for 2022 may be updated and replaced with 2023 values, and the content of the links overwritten as the data "moves right" in the yearly columns.

To solve this problem, one could manually remove each link, but that is undesirable, both because one loses the historical traceability and because of the labor involved in removing a large number of links.

Some partial, possible solutions might be:

(A) An ability to "freeze links" or "pause updates" from the destination location? Perhaps with a banner displayed, similar to what one sees when one has view-only access to a document, alerting all quite clearly that links are not being updated?

(B) An ability to create "pull" links as a special type of link, which would work more like connections, and instead of allowing link content to be pushed/published from the source, would rather allow the destination to pull/refresh it. (There's probably a good reason, but I still can't wrap my head around the rationale for why links, which are published from the source and automatically updated in the destination, behave different in this regard from both connections and published data sets, which must be deliberately refreshed at the destination.)*

(C) At minimum, an ability to "break links" i.e., remove all destination links en masse, so that the final result stays unchanged.

*P.S. For that matter, it would be nice to have not only "push" and "pull" links, but also truly synched links where the data is updated instantly and there is no publication or refresh necessary.