This is exactly the case in our institution (Museum für Naturkunde Berlin). Numerous artifacts like the mentioned correspondence, field notebooks and others exist and have on the one hand the role of information resources about the biological / geological specimens but are on the other hand administered as collections / aggregated resources in their own right. This is not only for administrative reasons (including storage needs). Many of these items are unique or have become so through, e.g., annotations of collectors or researchers which turn them into items with a museological value that contextualize the specimens they refer to in one way or another but are also informative for various fields such as history of science and various other cultural aspects.
Accepting a broad conception of what constitutes a collection in the context of the envisioned catalogue and the immediate connection these “grey areas” have to “proper” collections, I believe these could be generally included, possibly with content profiles of their own, as suggested in this post.
With regard to wether the catalogue should aim to represent deeper and more intricate relations among collections, such as above - it would be a good thing to choose a technological design which allows such representations to be added, possibly in a subsequent version of the catalogue, once the core structures are implemented.