One answer to that question is...that the obsession is partially an offshoot of another habit to add to the list. But this one is undoubtedly more widespread to projects outside software: estimating prospective risk. For planning any ambitious project, it's essential to start as early as possible to uncover and compensate for risks to success. Fantasizing about the advantages it will have after completion is allowable, but sooner rather than later someone needs to confront tedious questions about the grounds for its potential existence. What will be the signs that it's done? What will it be made from? Who will make it? How many weeks will it take them? How many overlooked mysteries remain about its behavior? In one sentence, what are the actual risks of actualizing it?
The habit of estimating prospective risk doesn't mean that the estimator is incapable of dreaming of improvements. It might mean that like me they grow...leery of suggestions which are massive yet ill-defined. The effort to take a risk inventory demands sharper clarity, because the larger or fainter it is, the more risk it carries. Until the team determines what it will be expected to do, they don't know the constraints they will be working to satisfy. And they can't productively speculate about the risks of achieving the constraints. Itemizing the conditions to make something physical depends upon sketching the contours of that something with high specificity. Whether the thing discussed is a project or not, whether it's awe-inspiring or not, the habit of estimating prospective risk urges the question of the tangible requirements it would need to fulfill for it to be real.