Ah, there are so many ways you could approach it. A few years back, the government messed with the idea of CAIV (Cost as an Independent Variable) where instead of designing a project and then estimating the cost, one would start with a fixed amount of resources (time and $) and see what you could get for it. There were some tricks to this of course since the development of software features is not purely modular, but that was an interesting exercise.

As for scheduling, you have to remember that while work may expand to fill a deadline that exceeds the minimum required time of a project, if you push a deadline that's short of the real cost for completing the project, you're still going to suffer a cost in quality if not outright failure.

I suppose in the end, one needs to gather the historical cost and schedule for projects past, and if you can gather enough data points, you can use it as an analogy for estimating future projects. Whether or not your organization tracks data to that detail is a whole different story.

That's my two cents as a labor estimator.