Simon raises an interesting point. I'm trying to understand this (so correct me if I'm wrong), but it seems to me:
If you have air in the system (e.g. the air at the top of the header tank) then unless the hoses etc. expand to an equivalent degree (which I doubt very much!) then the system will pressurise with heat expansion of both the air and, to a lesser extent, the coolant.
If you have a pinprick hole in a pipe, then coolant will be expelled by the pressure (unless air is present at the point of pinprick), decreasing the total volume of coolant+air. The expulsion will only continue until expansion of air+coolant has stopped
When the engine is stopped, the system cools and the air+coolant volume shrinks. Consequently a vacuum is created. As long as the pinhole leak does not have a 'valving' effect (i.e. as long as it allows flow happily in either direction - which is never certain) then air will be drawn into the system.
The cycle will repeat itself (only more exaggerated as air volume in the system is increasing each time) each time the engine repeats the hot / cold cycle).
If the air taken in at the pinprick point is able to get round the system, it should end up in the header tank, displacing coolant and causing a drop in level in the tank which, if noticed, allows a coolant top-up to be done.
If the air is taken in somewhere where it remains trapped, then the effects on the cooling system are unpredictable. You might get cavitation and loss of efficiency in the water pump due to air being present in that area. You may get hot spots in the engine. Who knows.
In the case of a pinprick way down the length of the bleed pipe, the air pulled in may just stay local, being expelled on heating and pulled in again on cooling.
Sorry if I'm completely repeating what's been said, but I'm trying to get my head around it because if I've understood correctly, there is just the remote possibility that pinhole leaks may produce unpredictable and hard to diagnose cooling system malfunction without drop in level in the header tank, although it may be hinted at by an unusual amount of rise and fall due to the pressure of hidden air in the system.
...and all of this before any boiling of coolant, and attendant vapour pressure, is considered
