There are many benefits to making a set of deliberately small - yet complete - examples of some art form.
Firstly, working within a challenging size-limit makes you focus on the core, or essence of each product. What does it really need to achieve, and having achieved it, what luxuries and polish are most worth adding?
There is no correct limit. It depends on what you're trying to learn and produce, on your current level of expertise, and available time. Setting yourself a limit which makes it difficult - but realistic - to create something worthwhile and then making a whole collection of things that meet those constraints will teach you a lot.
This also means finishing several products which brings us to the second valuable aspect: making complete, usable items.
It's easy to get stuck on a single stage of the production process, which we might roughly summarise as: conceive, create, refine, and release. Making a handful of complete products will often teach you more than starting a thousand and finishing none. Each stage offers different lessons and every stage informs the others.
Thirdly, you gain by making the items as a set, not merely a series. For example: start with an even smaller, rough version of ten items. Then make an improved, almost complete version of the best five. Finally, make a third pass to finish and polish. In this way every project in the final collection is informed and enhanced by all of the others. With each round you learn more about the general art these items are examples of, and your education in this art will improve the examples.
Of course, the fact that this is a valuable exercise does not mean it's the only valuable exercise. Far from it. But even if your aim is - for example - to make a single, much bigger project with no specific size limit, the exercise described here might actually save you time in that endeavour.
Making a practice set of smaller items first will let you test many of the principles, techniques, and effects you hope to include in the bigger project. This provides a relatively fast way to check a lot of assumptions, refine a lot of concepts, and stumble across many related ideas you might otherwise have missed.