We've discussed this a lot; some discussions have been more civil than others (some substantially less). Most people who actually chime in on the topic agree with your assessment. The current system doesn't model the way brewday actually works. And it always does new users heads in.
Brad has never really spoken up and defended "efficiency into the fermenter". Nor has he agreed to make any change based on our collective opinions. Other than acknowledge there is a bug in the way that extract brews are handled (specifically), he has been mute on the topic. I've tried several times to get his position.
There's no real good workaround, either. You can set your kettle losses to zero, that helps a lot to narrow the gap. But, if you have some funky mash-tun configuration (large dead spaces, or a tall false bottom---like the bucket in bucket config), you can still run into issues.
I've simply resorted to setting my brewhouse efficiency to about 5% less than worst case actual (for OGs below about 1080, above that I would drop it another 10% or so, maybe 5). I buy a little more grain than I really need to, and I typically stop lautering before I have a full kettle volume and top up with water.
In practice this actually works well. I never worry about undershooting my pre-boil gravity, or overshooting my pre-boil volume. I stop when I've collected enough gravity points, and top up with water before starting the boil. I can move pretty quickly through a brewday, because I'm not "tweaking" very much. I complete almost all brews in less than 4 hours this way.
The only downside is that I use an extra 1 1/2 lbs of grain or so over absolutely optimal. But, at $1.50 / lbs, so what?
Don't get me wrong, though. I still rather it were the other way around. I like accurate record keeping and I record and calculate a TON of stuff on brewday. I use javascript in custom reports to calculate the things that BS2 doesn't. I manually calculate my actual mash efficiency and record it on paper with my brewsheets. These get logged in a big binder organized by BJCP section, along with the BJCP guidelines. I've resorted to these practices because of this and several other shortcomings of BS2. I'd rather record it all within BS2, and have the paper records as backup, instead of primary.
