From the freeze frame data looks like bank 1 is running richer than bank 2 based on short term fuel trims. Not much else to glean off the data though.
Still could be problem unique to cylinder 3 - potentially a more global problem. I would have to go back to re-read what has been done so far, but, there is a possibility that the bank 1 upstream O2 sensor is failing. That could result in an inaccurate/incorrect fuel trim for that bank. It would cause problems more wide spread than just cylinder 3 though. Hard to tell based on information so far. We don't know what the bank 1 sensor reading is, but if we infer that because of the fuel trim for that bank being leaned out the sensor must be experiencing a rich condition. For that to happen there has to be too little O2 in the exhaust. If the sensor has some carbon build up, it could not read all the O2 that is actually present in the exhaust stream.
If you're the gambling sort, and running out of things to bet on you could throw one on it. It might come down to spending money on better diagnostic tools, such as one that can record a log of the whole trip, or spending money somewhat guessing at parts.
The alternative to that is that the sensor is reading correctly and is seeing the rich condition because a cylinder is miss firing and letting a bunch of unburnt fuel into the exhaust stream. We know a miss fire is occurring because at least we have a code telling us so and you can feel them. Still inferring from the short term fuel trim that the miss fires are on bank 1, it doesn't look like a global problem. It could still be just cylinder 3 or maybe another cylinder on that bank too.
Someone would either have to go back through all the items we've already talked about to confirm they are good, or start a more intensive diagnosis; in no particular order of magnitude, maybe smoke the intake to confirm no intake leaks, pop valve covers and inspect valve train, inspect cam/lifters/pushrods, borescope to look in the combustion chamber for anomalies.