There's two ways of benchmarking batteries: filling it to full and timing how long it takes to fail, and trusting/verifying what ACPI report.
I suspect you're asking for the former. What you might do is grab a clock, turn off battery settings and let it drain under some workload. When it fails, stop the clock and there you go.
GNOME does the latter, you can bring it up under battery properties. They keep a record of ACPI reported charge and battery life and the rate at which this changes during discharge time. What you end up with is a good prediction of history in the middle, with less confidence of the prediction at the end because you don't often run a laptop to dead, and when you do, disk writes become dicey. The end near 100 percent decline is also unreliable, for reasons I'll explain next.
One caveat to modern laptop batteries: their performance degrades over time. My laptop battery that lasted 4 hours when new lasts 30 minutes at best now. I'd be interested to see anyone who has data on how a battery declines over time and use. Every time your battery charges to full, it goes to a slightly lower voltage / capacity, and that's why you don't have good confidence at the top end; you visit it every time you unplug and every time is different.
But let me caution you against the drain-to-failure approach. What you really care about is how it will perform in some future situation. Will you be able to watch two movies on your flight? Or take notes in this meeting, etc. In that situation, you don't care about the extra 1 percent you might get by skipping hibernate at critical battery life; you care about the integrity of your work. Just take the system off AC, do your job, let it hibernate when it normally would and check the clock.