Bring me my beverage of choice!
With the Aeroforce gauges that I have, it doesn't ask for the diameter, but the rolling circumference in mm. So what I did was make a chalk mark & roll the car one rotation, then measure. Worked out pretty close.
As for speedo accuracy, the specs used to say +/-10% at or above 40kph, but now are a little more complex,
from ADR18/03.:
(JK is type M, my bike is type L3)
V1 is the indicated speed
V2 is the actual speed
In the case of vehicles of categories M and N:
0 ≤ (V1 – V2) ≤ 0.1V2 + 6 km/h;
In the case of vehicles of categories L3, L4 and L5:
0 ≤ (V17 – V2) ≤ 0.1 V2 + 8 km/h;
So from this, at 100km/h actual, your speedo can indicate anywhere between 100kph, (must not under read), and 116kph. (100 * 0.1) + 6
My bike speedo can indicate between 100 & 118kph.
So Joe Bloggs can think he is driving down the highhway at the posted 110kph but is only actually doing 94.5kph! And you wonder why so many people seem to be driving so slowly & swearing at the 'speeders' who actually just have accurate speedos?
For the record, at 110 actual, the speedo on my bike reads 120kph. And the odometer is accurate with this setting. The JK, on the gripping hand, has an accurate odometer when the speedo is accurate.
One point, though, Seacomms. The pressure still makes a difference even to a bald tyre's rolling diameter, due to the pressure dependent change in deformation. At high pressure the centre of the tread is pushed out further than the edges but as the tyre pressure drops the centre of the tread begins to bow inwards at the contact patch. (This is why an under inflated tyre wears more on the edges than in the middle, whereas an over inflated one wears in the middle of the tread first.)