The air line blowing off the diffuser is obviously a sign of pressure finding the weakest point of escape. To get any definitive answers, you must eliminate excess variables when testing. If you are questioning the diffuser, make it your constant and remove all variables.
First, remove any question as to the quality of the tubing-to-diffuser connection. Heat the tubing in very hot water to make it pliable, and then let it cool once you have placed it over the diffuser. This heating and cooling cycle will cause most tubing to shrink, ensuring a tight fit.
Now that you know you have a good connection, try running the diffuser by itself, straight off the Co2 cylinder. If it works properly, you can rule out the possibility of the diffuser being defective. Make sure to check your secondary regulator pressure gauge to be sure that your regulator is providing a reasonable amount of Co2 pressure.
I can tell you that it's hard to dial in needle valves on multiple Co2 distributions. In your case, you have three different distribution methods, and each method has it's own pressure resistance.
I gave up on trying to power multiple Co2 distributions off a single cylinder. For me, it was worth it to purchase multiple cylinders. That may not be an option for you, but if you can afford multiple, smaller cylinders, the benefit of individual stable systems can easily be worth the investment in the long run. You will spend less time troubleshooting, have more confidence in your systems, and possibly fewer trips to get cylinders refilled (assuming all three tanks are roughly the same size and have roughly the same plant/light/nutrient conditions, all three cylinders should run out at roughly the same time)
If getting two more cylinders is out of the question, and if the diffuser turns out to be ok, the only advice I can give you is to continue fiddling with your needle valves. You should understand that the more distributions you add to a single cylinder system, the larger the impact of any individual needle valve drift. A slight drift in one needle valve may completely shut down one or more of the other distributions. I could not deal with that kind of volatility in my own personal experience.
"Insanity: doing [or asking] the same thing over and over again and expecting different results."