Good question. To extend on what was shared above…
Most times I’ve had no problem with boot order. I have had more fussy behavior with busses that are extended to the limits of the pull-up resistance. Due to the number of devices and amount of cabling connected, they can be in a working - but in a brittle state.
The first place where you can see things start to get weird on a bus is with i2c reads. When the bus is weak - they start failing and can cause the sending device to lock up (in the case of the Teletype). It may happen very rarely, which can be difficult to isolate and troubleshoot.
This can be problematic if you have a script on the Teletype starting up immediately and firing off commands - or a 16n that starts broadcasting i2c the moment it is turned on. On an unstable bus, or a bus that is still getting up to the proper pull-up resistance to be stable, you can have problems.
I do have a crazy setup on my dining table right now that is bridging 3 micro-cases with 9 i2c modules and an external 16n. With it, I’m finding that boot order matters. Turn things on in the wrong order and the whole thing hangs. It is consistent with @okyeron’s point above - followers first - then the leader.
That said, I’ve done crazy stuff with i2c and have had lots of devices connected without problems. One you have a stable bus - you usually have a stable bus. But, don’t be afraid to experiment with cable lengths, device spacing, and pull-up resistance to make sure you have a stable configuration.
Now - specifically on the ER-301. There was a time where there was some weirdness on startup. Not the exact situation, but it was startup related. This was corrected some time ago. Here is a reference to that area of the thread over on OD if you/anyone are curious.