Well, that was fun,... and useless. I took a miniature fire hose nozzle which fits on a garden hose and measured flow rate when it was in the water in the bucket and out of the bucket. No difference. I measured reaction force when the tip was at various depths in the water and when it was out of the water. No difference. It surprised me that it was only 20 ounces (1.25 lb) in all test situations. Felt like more. For reference, the pressure at the nozzle was 30 psi when the nozzle was flowing (there was a long section of hose ahead of it).
Flow rate was 1150 cubic inches per minute into the 3 gallon bucket. Nozzle exit diameter is .190 inches. So the exit velocity was v = 1150/[(pi/4)*.19^2] = 40560 inches per minute = 676 inch/sec mass flow rate was m dot = [(1150/1728)*62.4]/60 = .69 lbm/sec (.69 lbm/sec)/(386 in/sec^2) = .0018 lb-sec/in reaction = m dot * delta v = .0018 * 676 = 1.22 lbf What have I proven? The jet reaction is real, and measures as predicted, but seems to have nothing to do with the Moss effect. The Moss effect must depend on the configuration of the spray coming out of the nozzle. Also shown was that in this case, the flow around the nozzle while it was in the bucket was not a part of this problem. Indeed, the water in the bucket looked completely turbulent. John B PS to the rational countries: The American system of units is very simple. Just throw in the acceleration of gravity (386 in/sec^2) wherever it is not needed and omit it wherever it is needed. Once you have that down, it is an easy system to use. My tools were American system so rather than converting...
