I'm sorry to say this is wrong. Regardless how much coolant is in the radiator, the coolant pressure created by the coolant at the sensor depends only on the height from the sensor to upper hose -- about 10". The pressure doesn't depend on if the radiator holds 50 gallons or 1liter of coolant:let's assume 1/2 of that is in the radiator. 4.4L of fluid, with the aid of gravity is trying to rush out the open lower radiator hose opening. You can do the math exercise, but we can calculate how much additional vacuum is needed to actually suck enough of that coolant back up the radiator in order to introduce air. Do you think the additional vacuum that one would introduce by squeezing the upper hose is going to overcome the head (liquid pressure) of the coolant in the radiator due to gravity? I think we know what the answer is.
Pressure = p*g*h; where p is liquid density, g is gravity constant, h is the height of liquid.
So imagine if we have a very long, a 10" long eye dropper with a massive E46 upper hole attached to the pipe top instead of the little rubber bulb. Now squeeze the upper hose as MrM instructed with all your weight, do you think this upper hose "bulb" will be able to suck in air when we let go the "bulb" to return to its glory shape? I say it will.
People might think the pressure at the bottom of a 10 feet pool is larger than the pressure at the bottom of a 10 feet drinking straw, but they are the same.