@Jens-Jensen Thanks for posting that video. I wonder if in Julien's' case there might have been moisture penetration through the insulated wiring ingress points. If the outer insulation coating were sufficient, then waterproof wiring wouldn't require that sticky goo they put underneath it. Once it gets inside the "sealed" enclosure by travelling through the insulation, it can do its damage with nothing to stop it. So....maybe if Julien had applied something like Corrosion X or some other conformal coating and then sealed it all up it might have lasted longer. I think extending the heatshrink to completely cover the wiring so that only the metal connection terminals on the end are exposed may be the only way to prevent the moisture ingress, assuming the heatshrink stuff really is moisture proof.
In contrast, I suppose in the Great Scott epoxy filled approach, even if moisture does ingress through the wiring insulation, it has nowhere to go since all the electronics are protected with the epoxy.
For those who don't know: something can be waterproof without being moisture proof. Housewrap would be a perfect example of that, as would gortex raincoats. So, having something that's waterproof rated doesn't tell you enough information. i.e. even something rated at ip68 could let in moisture. A lot of plastics are waterproof, but not moistureproof. Also, from what I've read, the most commonly used hot glues are not moistureproof.
And of course if you have any air at all inside, the moisture in it might condense into water if it gets cold enough.... So, I think Julien's setup is prone to eventual failure for a whole host of reasons in addition to those that he mentioned: the host glue, the wire insulation travel path, and the trapped air.
I did once try an experiment using just hot glue to seal cheap corrosion prone chinese electronics (pretty much engulfing it in hot glue as Julien tries at the end of his video), and I was surprised that it failed in just a month or two of wet outdoor weather.
