The more I read about the differences in the WiFi 6 standard, the more dubious I am about certain aspects of it for my purposes.  First, I have 3 Pro gen 2 nodes.  I decided I could use one more node, and I have a need to buy a separate 2 node network for a different location, so... (also, please think in terms of what I am asking and don't focus on the iPhone 12 client, which is just for example).

I am thinking of buying 2 Eero 6 Pro nodes and re-purposing 2 of the 2nd gen Pro nodes for the other location.  The re-purposed ones would be overkill, but buying 2 Eero 6 Pro nodes for the other location would be criminal overkill.  I could probably do with just a regular 2 node Eero dual band set up for location B.

My original thinking is to:

1) Replace my Pro 2nd Gen gateway node with a Pro 6

2) I would add a Pro 6 as the 4th node

3) I would keep 2 Pro 2nd gen's for a total of 4 nodes (gateway 6 / plus new 6 / plus 2 of the original gen 2 nodes) for a total of 4 mixed ac/ax tri-band nodes

Now - I assume when I am directly connected to my gateway node with a WiFi 6 device like an iPhone 12, I probably have full benefit.  My question is what happens to the 2nd Eero 6 Pro node?  Does it just default to the lowest common denominator because it's part of a WiFi ac mesh?  In other words, if I connect to it with an ax device, I assume it's connection back to the gateway is still going to default to ac because the other 2 nodes are ac?  Lastly, can my gateway Eero 6 Pro even treat an iPhone 12 as an ax device because it's already serving with a mix of ac and ax nodes?  Or would it just switch between ac and ax to that device as it moves around the house?

First tier support is great, but they're not network engineers, and nobody so far really seems to know for sure any of these answers