[serialk11r]

Sorry that was perhaps a slight exaggeration. You don't really need a BSFC map though, it's very reasonable to expect that the low load efficiency is going to be similar to say a 2ZR-FXE, which has similar cam timing, although direct injection changes things a bit.

A Prius can cruise at over 95% efficiency, while a Honda S2000 cruises at around 60% efficiency. That's the kind of spread you see even among small engines. That's the difference between good gearing and bad gearing really.

The AT will require about 52Nm from the engine at 60mph, while the manual requires about 40Nm. The 2ZR is 1.8L instead of 2L so we can multiply these by 0.9, which will give the approximate corresponding load on the 2ZR.

The AT will be able to run the engine at around 270g/kWh efficiency. The MT runs the engine at around 300g/kWh efficiency. Peak efficiency is slightly better than 230g/kWh.
Sorry for all the edits by the way.
EDIT: I'll add in the numbers at 75mph for reference, in a moment.

Okay 75mph, AT needs about 77Nm from the engine, the MT needs about 60Nm. AT is running at 240g/kWh, MT is running at a bit over 260g/kWh. And at 80 (do 80mph speed limits even exist? I haven't seen one before...highest I've seen is 75, so that's technically the fastest you should drive :P) MT is a little over 250g/kWh, while AT is at 230g/kWh. So consistent ~10% difference or so.

Also note that at 80 your engine is running pretty efficiently, although the power requirement is much higher. With a longer rear diff or slightly longer 6th gear we can close the gap pretty much all the way, especially since an AT packs hydraulic pump losses, which eat up part of the advantage.

Moral of the story is, if you care about mpg, in the city, go manual. Lots of highway miles, AT will have probably somewhere between 5-10% better fuel economy on average. EPA test is done at lower speeds, at higher speeds the gap will close up a bit. The reason being, a modern engine has a huge range in which it is within 10% or so of peak efficiency from 100% load down to the high 30s, and across several thousand rpm. Efficiency drops off rapidly in the area where the EPA test happens, cruising at low speed, and the MT will get to the low efficiency zone faster.