There's no straight easy answer, without knowing all the details of the compared systems and the testing done to compare them, including test conditions.
If the drive they compare to is really not very efficient, then their drive could be that much more efficient.
Being brushless vs brushed is more efficient, so ther'es that.
Also there are different magnetic designs of motors, and while you can just punch out a bunch of laminations, stack them on a shaft, and wire them up, that doesn't mean it works as well as it could if you optimized it's design. Could be the difference between 60-70% efficiency and 98%+ efficiency, at the right RPM/voltage/load.
A motor that runs closer to it's unloaded speed, when loaded, will be more efficient (less current to drive it) than the same motor running much less than it's unloaded speed, when loaded the same.
Since the motors arent' the same, they can't be directly compared, but it is possible that the DD brushed motor is running far enough below it's unloaded speed at the test loading that it is far from it's intended efficiency, and that the geared brushless motor is running close enough to it's unloaded speed at the test loading that it is close to it's intended efficiency, and the difference could be huge.
Gearing a motor down means you can run it at higher RPM under load and keep it closer to it's efficient zone, than with a directly-driving motor, all other conditions the same. So it's possible that being a geared motor makes a significant efficiency difference---even though the gearing itself will eat some of the efficiency (a few percent, typically).
It depends on the test conditions, and if they are biased toward one system or the other, or if the conditions are even the same between the two tests (some marketing departments provide the worst case test conditions for their competition, and best case conditions for their own stuff). If they provide the test condition data along with the test results, then you can judge for yourself.
Otherwise, there's no certain way to know without testing both yourself under the same conditions.