For moderate subject distances, the total depth of field (near to far) is proportional to fnR**2 where
R=reduction (= 1/magnification)
For a subject of a given size (say a head and shoulders portrait) the reduction R is proportional to 1/L where
L = sensor dimension
Since we are only interested in a ratio, we can take any consistent sensor dimension in the comparison – diagonal, horizontal side, or vertical side. Choose 11 mm diagonal for 2/3" sensor, 22 mm diagonal for 4/3" sensor.
Comparing the total depth between the X10 and a M4/3 camera shows it is a wash, using f2.8 for the X10, and f5.6 for a kit lens on an Olympus Pen E-P2, say, using the ratio of crop factors for the ration of focal lengths.
Terry's calculations essentially agree with this.