The problem with the altimeter setting values are that they are "Manometric" units - used for manometric calibration - and because they are "derived units" there is no proper international standard for them.
The sea level conditions for the International Standard Atmosphere are 101.3 kPa at a temperature of 388.2 degrees Kelvin (15degrees Celsius) and a density of 1.225 Kg/m3. The sea level conditions for the American Standard Atmosphere 1976 are 2116.7 lbs/ft2 at a temperature of 59 degrees fahrenheit and a density of 0.002378 slug/ft3 which amounts to the same thing but in US units.
That 29.92 inches comes from the meterologists standard atmosphere however, by which the sea level conditions are taken as 29.92 inches of mercury and a temperature of zero degrees celsius.
Being derived units, derived from different standards, there is thus no direct equivalence between the 29.92 inches and the 1013.25 millibars of the two scales. It remains correct that for practical purposes the difference is too small to be of any concern to airmen.
It does matter to us instrument calibrator types though. Small differences at lower altitudes are magnified logarithmically as one ascends the scale, leading to much frustration for the poor calibrator who is tweaking them out by trial and error. A further complication comes from older altimeters being calibrated to older national standards that differ in their base values. For example most altimeters of British construction dating back prior to 1976 were calibrated to 'London Laboratory Conditions' - 30 inches of mercury and 60 degrees fahrenheit if my memory serves me right. It is too many years since I applied Fiducial Correction to my Casella Barometer readings using the trusty old "Z" chart...