<... Why wouldn't the pressure increase when water levels drop and steam develops? ...>
Need to work backwards on this one.  If the water level drops, one would deduce that it happens for a reason, and a leak comes immediately to mind.  When the cooling system starts leaking, the pressure drops off substantially quite a while before any significant drop water level.  FIRST the pressure drops precipitously, and THEN the coolant level starts to go down ... FINALLY the temps start to rise (provided the temp sender is submerged in what water remains).

If you were to use a coolant level indicator, where would you put it?  Not in the block, because Ed has established that what coolant remains is relatively well distributed in the block.  In the expansion tank?  I would suspect you'd get a lot of "false negatives".  I don't know where I'd put one, or how much to trust it.

Just a theory .... Jim S.
 

Finn Lassen wrote:

Ed Anderson wrote:

>It is my opinion that a coolant pressure gauge is a better early indications
>of a leak rather than coolant temperature.  By the time your coolant temps
>start to climb (even if the temp gauge is reporting the condition accuracy)
>you are already in sufficient trouble that probably only turning the engine
>off and becoming a glider is going to save the engine from damage. ....

 
Why wouldn't the pressure increase when water levels drop and steam
develops?

I still think the best leak detector would be a water level sensor.
Petty that the solid state one for fuel tanks is not designed to handle
higher temps:
http://www.ppavionics.com/LFL.htm

Here's a mechanical that Rusty found earlier:

http://www.grainger.com/Grainger/productdetail.jsp?xi=xi&ItemId=1611587508

Finn