Also, I think you're confusing things. Are you referring to this?
http://anandtech.com/show/7085/the-2013-macbook-air-review-1...
That shows 11 hours in "light workload". Unfortunately, starting with this year, Apple stopped giving "real" numbers for its Macbook battery life, and started giving battery life for "minimum workloads" like the whole laptop industry has done for more than a decade. The "real" battery life is closer to that "medium workload" (Haswell gave about a 50 percent boost compared to last year's Macbooks).
With medium workload, it's 9 hours, with a 50Whr battery (according to you, I didn't check), so at 30Whr, that would normalize it at around 6.7 hours. Still good, but nowhere as impressive as you made it sound at first.
I'll grant you this. The decision to use a last-gen chip in this Chromebook wasn't a very smart one, and that's probably HP's fault. Mediocre companies like HP with mediocre employees and mediocre middle management tend to be very slow about adopting new technologies in their products, which is why you never see them come out with some cutting edge tech first.
With Haswell Intel got the power draw at idle down to a very good level. So the major power sucker is still the screen. But given the Chromebook has a smaller screen and less powerful SoC the ChromeBook should in theory get little better battery life than it does. That one we can chalk up to the 3.4 Linux kernel that the ChromeBook is using and OS X / Safari generally always being ahead in power use department.
I think we'll see the battery life get better with the newer Haswell Chrome books when Google switches to a more recent Linux kernel.
To be accurate:
The Chromebook gets 38.5% of the battery life with 55.5% of the battery capacity. That's quite a bit different than the 1:1 ratio you imply, more like 0.7:1.
Put yet another way, with the MBA13's battery, the Chromebook would last 9.7 hours. With the Chromebook's batter, the MBA would last 7.8 hours.
If that's not implying a 1:1 ratio I don't know what is.
I'm betting it's Chrome OS. The Nexus 10 also runs the Exynos 5, same exact chip, yet it gets 9+ hours of web browsing battery life and does it with a 2560x1600 screen.
In this case, it's common sense. There is no way a Nexus 10 has a battery four times as large as an 11.6" notebook.
Chromebook has a bigger battery than the N10, not a smaller one. Probably also a factor in why the Chromebook tips the scales at 2.3lbs vs. the N10's 1.3lbs.
"Despite using relatively power efficient hardware and being paired with a 30Wh internal battery, the Chromebook 11 barely lasted 5.4 hours in our web browsing battery life test. Local video playback was even worse at 4.8 hours."
The "web browsing battery life test" in this article is the same one as in here: http://www.anandtech.com/show/7117/haswell-ult-investigation. Note how the iPad 4 gets 9.48 hours on both tests. On that test, the 11" MBA gets over 11 hours, despite having less than a 25% larger battery and using Core i5 processors. The 13" MBA gets over 14 hours, though it has an 80% larger battery (but has to drive a 30% larger screen).
This suggests that Samsung's Exynos 5 is extremely power-hungry or Chrome OS's power optimization is extremely bad.
As an aside, Apple/Intel's phenomenal battery life performance with the latest MBA signals to me the beginning of the end for ARM. If we're getting to the point where a current-generation Haswell laptop CPU is putting up light workload power draw comparable to a flagship A15 tablet chip, that's not a positive thing for ARM. Heck, even if you just compare the iPad 4 with the MBA 11", the writing seems to be on the wall. The iPad has a substantially bigger battery than the MBA, while driving a smaller (albeit more pixel-dense and power hungry) screen. And yet the MBA totally holds its own on the battery life test.