Doesn't using a one-hex-per-state model basically defeat most of the benefits of using hexagons in the first place? This seems like a bit of cargo-culting the idea that hexagons are good for maps without actually understanding why.
They look neat, and that's probably a sufficient reason for most people to use it. For anyone looking at optimizing some problem for which hex grids are a solution, they're not likely to be using excel, I think.
I think the biggest question is can hex grid approximate US state topology/geography better than square grid, which would be the main alternative. Intuitively I'd guess yes, but would be nice to verify that.
I tentatively agree, I'd try harder to keep coastal states on the "coasts" in such tile map. That being said, I don't know how hard that would be in practice.
But majority of states have borders with >= 4 neighboring states and on top of that you have coastlines, so that leads me to think hexes might work better.
I wish there was more medium level content like this available for office.
It's easy to find basic tutorials of the 'Press the B button for bold' level. You know, where messing around with the interface for a few minutes brings you up to speed.
Also easy to google for VB stuff and manual-style things. What function or code sample do you need to accomplish X.
But the level in between, where you find about less known functions and interesting usage of tools, is hard to learn. You can stumble around until you find a gem, but I am not aware of organized repositories.
One of the last steps is to translate all the odd rows horizontally by the same amount, so that the hexes line up. That makes the correspondence between what you drew to begin with and what you end up with strange. I'd find a linear transformation more intuitive.
https://www.redblobgames.com/grids/hexagons/