Without anything there in place of the t-stat, the liquid will run to fast through the radiator and not have time to cool.
Not to start a stir, but although often stated, this is not correct.
Although a period of time is required for the transfer of heat from the coolant flow volume to the inner surfaces of the heat exchanger, there is more than ample time for this in most automotive executions with or without the thermostat restriction. Remember, if you slow the coolant down in the radiator, you've also slowed it down in the engine, therefore since the engine is in a constant heat production state then the offset to the theory of the water experiencing a greater temperature reduction upon leaving the radiator is that the water is going to have a greater heat gain upon entering the radiator, with the reverse with the coolant moving faster, within reason of course.
The function we are referring to here actually is the "restriction" being at the discharge from the engine makes the plumbing from this point back to the discharge side of the impeller the "hi-pressure side of the system, of variable pressure in excess of the cooling systems general "pressure rating" accepted as the pressure rating of the pressure cap which is the low pressure side of the system. Since the orifice restriction area remains relatively constant (particularly with simple restrictor) but the flow rate is variable, increasing with engine speed (R.P.M.) the intended result is an increased pressure sum with the increase of load/work done, as there is greater heat production, this increased pressure further resists the potential of localized boiling or hot spots particularly prevalent within the cylinder heads, simply by raising the coolants' boiling point.
Now I've started trouble!
Scott.