How tight should the PTFE runs be on an INDX setup?

Leave the slack. Shortening them is the one thing that'll acutally cause a failure, because a tube that's too tight pulls straight out of the ECAS fitting the moment the head reaches the far corner while the spool side is also tugging the other way. The bowing you're seeing on the outer channels is normal, it looks worse than it is. What matters is whether the tube ever goes taut, and a full travel pass is exactly the right test. I did the same thing after my build, ran the head to all four corners and both extremes of the gantry by hand with the steppers off, then watched a couple of full range moves. If nothing snagged and the tubes stayed curved rather than straightening out, you're fine.

The thing to double check on that far right channel is what happens when the head moves toward it AND up at the same time, or when a toolchange is happening. Toolchanges are where people get burned, because the tube that's retracting while another is feeding puts a diferent load on the runs than plain head movement. Do a few manual tool changes if you can and watch the fittings at both ends. As long as you've got a visible arc the whole time, you've got enough. On zip tying pairs, I stopped doing it entirely. With eight runs coming off a selector, the tubes move different amounts depending on wich tool is active and where the head is, so bundling any two together just makes one drag the other. Each channel needs its own little amount of freedom. If the bundle looks messy, use a loose guide clip near the top where the movement is smallest, but leave the lower runs separate. One more thing, make sure the tubes have a bit of slack at the selector end too, not just at the head. Mine were fine at the print head but tight coming off the rack, and that's where I saw a fitting pop once. Better to find that now than mid print.

Testing with filament loaded matters more than I expected. My first travel passes were fine, then a week later a tube started lifting at the fitting and I realized dry PTFE bows one way and loaded PTFE, especally with a full spool pulling from behind, behaves completley different. Redo the homed travel pass with all eight feeds primed and see if that far right channel still has its arc. Changes things more than you'd think.

Mine settled in after a week or so, the PTFE kind of relaxes and the arcs get looser on thier own. So don't panic if it looks tight today, just recheck everthing after a few prints before you trust it on a long job. I also marked my fittings with a sharpie line so I coud see at a glance if one was walking out. On the pairing thing, I only bundle near the very top where they barely move, everything below hangs free. Sounds like you paired them too far down, that's usally why the short one yanks the long one around.