I've since changed it quite a bit and it seems to be a fair deal happier. This is the new table I'm running at the moment:
It's more like the EFISource table with the idle position adjusted to my car and the bins changed around a little bit. It's way happier on this one and will cruise on the highway fine like it used to. Also has some more zip to it for sure, though I still think it's likely got more to give if I want to push it. I also got away from using the two different tables for the time being. My initial setup used something similar to the factory with an "open loop" and "closed loop" table, though really for me it was more <80% TPS and >80% TPS since with a wideband I'm kind of "closed loop" all the time. That's the main reason there is the big hole in the upper right. If I had the dual map setup going this table would be smooth all the way across and the second table would be where the timing is taken out for hard throttle hits, but for the time being I figured it was easier to tune a single table. Now I cruise on the highway around 27 degrees of timing and it seems like it likes it well enough.
I don't really like the huge hole at idle, though I'm not sure if it really matters or not. I can actually set up an idle advance table in the Megasquirt so that the engine swaps to a much smaller table (it's only 4x4 or something like that) whenever it's in the "idle" state, but I'm not sure if it would actually work or not. The main reason I think it might not work well is that if I take the hole out of this map and bump all the timing back up so it's a smooth curve, the engine will idle really high (it's basically like 20 degrees of advance at idle), and if it's idling high I don't think it will ever drop into the closed loop idle. I may still give it a try someday, but for now this is running well enough. I plan to try to get the knock sensors working as soon as I can get my hands on a knock module, and at that point I think I'll probably see about bumping up the WOT timing numbers.
The other thing I've thought about doing is reverse engineering the timing map in my parents' Ram. It's a 2016 or something like that, so it's not apples to apples since it has the 09+ heads and the VVT, but I'm thinking it would still be interesting to see and compare against. I have a bluetooth dongle and should be able to datalog with the Torque app on my phone, so I could try to rebuild the map. The more I think about the "factory" map that I had, the more I'm not convinced it was really the final timing number at any given point. I feel like there might have been other modifiers affecting the final timing value and that the map I had was just a base that gets added or subtracted from. Datalogging realtime would let me see actual commanded values at any operating point.