I've been doing a lot of pitch work recently - giving big ideas and high-level overviews of proposed new projects. It's a fun time in a project because it's incredibly easy in some respects - since the project does not yet exist, it's trivial to make adjustments.
Cutting Features, Adding Features, Feature Creep
Lots of my colleagues speak in terms of "features" for each new game. Each new feature is geven a cost (in time, effort, and money) and groups of features are added, removed, and traded in these early meetings. But something about this language hasn't been sitting well with me - because I know from being on the production end that "features" are not modular elements we can manipulate individually - no more than you can simply remove a wall from the side of your house or cut out a middle floor from an office tower.
Instead, I like to think of my project proposals as being built out of Playh-Doh. The question is "how much do we have?" And then we simply allocate to fit our requirements. You can make nearly identical things out of differetn amounts of Play-Doh - the only difference is their scale. If you can make a castle out of X amount, you can make a smaller castle out of Y. You don't build half a castle - you build a whole castle which is smaller.
Ruled by Constraints
Of course there are limits to how small you can make things - we call these constraints. If you have some doll furniture which needs to fit inside your castle, that provides a constraint. If you need to be able to stand inside it - that's a constraint. So it's not the case that projects made with less material are identical scaled-down models of larger proposals - but they are still whole-unto-themselves entities.