I can't find any trouble from your photos or description. It's unlikely that the RGB led is causing trouble, it works with a different protocol (WS2812) and it's data line is completely isolated from the Mux.
At this point you should concentrate on getting a 0x73 response with i2cdetect and no K-Ceptor plugged. The lines between the RPi and the Mux are exclusively in the area btw TP11 and U1. You can see them passing through TP8 (SDA) and TP9 (SCL). If you have access to an oscilloscope or logic analyzer a goo idea would be to test the signal using those TP.
(By the way, in the picture I see a blob of something in the proximity of the SCL pad of U1, if it's solder paste then it might be bridging the two bias on the proximity.. unlikely to be the problem, I know, just guessing)
In order to eliminate any other source of troubles.. have you test the i2c communication from your RPi with another device?
If none of the above applies, the only thing I can think of right now is a broken PCA9548 chip, or perhaps an unsoldered pin. Try inspecting them closely with a microscope.
