Best way to start appreciating the things Handbrake does, is to believe what the developers (who are professionals that volunteer their time) say.
They are trying to juggle several discontinuous libs into a manageable package, and have said over and over that the updates to those libs and the shift to quality-leveled rather than bitrate-leveled encoding makes that minor feature (which you can calculate yourself in a simple excel file) impractical to chase through every incarnation of the software.
TedJ wrote:You'd be amazed how many people would register on the forums to complain of a 1.5% error in their desired target size.
Actually the simple formula I gave is only accurate within about 5% at predicting file size in the real world of long-GOP interframe compression, maybe even a bit less accurate with some high profile compression schemes. As usual, I ran my own tests. I'm sure the developers here are less interested in writing predictive regression algorithms than in improving the software itself, and understandably so.
But if you are convinced it is as easy as you say, ejtttje, I (and many others here) eagerly await your code submission.;?)