But we are not considering three 'points', we are concerned with the surface areas of 6 pairs of flat thrust faces (2 pairs per isolastic), ..
Well.. it took a while , but I was expecting this to turn up .
The “ 3 points , one plane” is of course a simplification .
In reality it is 4 parallel planes , or 6 , depending on type of headsteady .
(a tie rod type headsteady is supposed to have no lateral play at all .)
But because of the relative small diameter of the mounting collars and a 0.2 mm gap , it would take an awfull amount of deformation for them to bind up .
While in a 4 point setup , a tiny movement of one of the points ( you suggested 0.001 “ ? ..) all 4 would bind and transmit vibration .
BTW ,the thrust faces of the collars are often not parallel , so I believe it is better to use a dial gauge instead of feelers to set the isos .
On the same page of this forum , there is a thread about cracked frame tubes .
Suggestion is that it might be caused by too tight isos .
I “ heared “ that bolting the engine solid in a Commando frame will cause the frame to crack .
Don’t know if this is a myth or a real risk .
If true , it would be a compelling argument against a 4th iso..