Just thoughts, maybe not a solution..... In my experience of working on many electronic systems over the years I have yet to run across a transformer that was running hot because it was bad or defective - almost always because of high or excessive load. I'm sure it happens, just that it seems rare to have the transformer being the cause. And it's not really unusual for them to give off a slight odor when running near maximum capacity and temperature. That said, people are buying / replacing that transformer, sold on amazon (for over $200, yikes), so perhaps something is going on. Also, I don't see any discoloration in your picture (even the label looks good) so still wondering if it's "too hot". I suppose the only scientific analysis would be to know the ratings for load and temp and compare to actual - a thermal IR gun or a meat thermometer could compare if you could find the rated max temperature. Or perhaps a call to Hayward would help. One simple test would be to shut off the SWG for a few hours and see if the transformer cools, then you would at least confirm whether the SWG is the culprit load (if there is a culprit or problem at all).
But you have another symptom: It seems very wrong to me that you might find a need to run that T-15 cell at 60% on a 12K gallon pool to achieve the needed chlorine level, even in the current warmer temps. It's nearly as warm here in SW FL with my 16K gallon pool rising to 89 on it's own (no heat running), and FC is always 2-3ppm with my aquarite T-15 cell set at 20% - even with our recent washout/deluge of 10" of rain over the last few days. I did have to add salt due to overflow loss. I usually run the T-15 at 10-15% in cooler weather. Keeping my CYA at 50-70ppm does assist that as well. Given a design life of the T-15 based on a 40,000 gallon pool, 60% tells me there is another issue. Not sure what that issue is - low or zero CYA, bad or dirty cell, defective controller, even the transformer? I don't have automation, just the stand alone Aquarite AQR15.
There are many threads here in TFP and elsewhere on the web that describe cell current draw expectations for various conditions, that may help. When testing it's also worth bearing in mind the active/rest cycles of the SWG based on percentage settings. It seems that my "generating chlorine" led is flashing even when it's in a rest cycle, so that can be deceptive. Those rest cycles were kindly described by @
BMerrill here as well:
SWG output: Discussion on Cell Cycle times vs. Scheduled run times
I hope some of that leads you to a positive place, and we're always interested in hearing from you about the ultimate resolution!