I’m guessing, but I’m guessing probably not. But maybe. It all depends.
i2c is fussy about cable lengths. It wasn’t designed to be as fault tolerant as something like USB. When the bus gets too long, things stop working reliably. There is no easy way to math it out that I’ve found given how wacky things get behind the panels with the busboards, individual modules, and interconnects.
What I know is that if I use a rather short stereo cable, I can connect two rows together using two TXb. I’ve also been able to connect a 16n via a short cable to that configuration and have it work. Your milage will vary based on the number of modules connected to the bus and the length of your cabling (and the resistance that your specific cabling introduces).
Anyone wanting to do a complex i2c arrangement should be prepared for some patient experimentation. The basic rule of thumb here is closer together is better - which means that less cabling is better.
I’ve personally managed some massive setups this way - like a Teletype connected to 8 TXi and 8 TXo.
Though, another thing that I do is implement a daisy-chain setup between modules as opposed to a star configuration. That is why the Telex have two sets of the i2c pins separated from each other on the back.
So you can use short interconnects like the little one on the left:

Sorry there isn’t a clearer answer.