Then take the deflector off for a flight and do the test again.

Back to back testing is one way to improve the odds you're doing good testing. BTW, my 49AXs all come back too hot to touch, and I'm absolutely positive they're not running too lean. And all have lengths of silicone tubing on the mufflers, even on the Pitts mufflers. But no lengths are over 3".

Two cycle engines usually do not overheat when a 2" or 3" flexible hose is added to the muffler's outlet. The commercial ones have rougly the same inner diameter as the mufflers they fit. And the silicone tubing that fits the mufflers has almost exactly the same I.D.

Any two cycle can overheat with too long an extension, but the overheating comes from back pressures that're greater than the design expects. And that backpressure would create a loss of power, because it would be stopping the hot exhaust gasses from exiting the cylinder. And those backed up gasses would be inhibiting the incoming fresh charge. So you would see a reduction of the engine's power. It would show up clearly on a tachometer. And any experienced modeler would detect the loss of rpm without a tach.

If you're not detecting an obvious loss of rpm, the conditions for overheating aren't coming from an increase in backpressure, which is the only thing the extension could do to cause overheating.