I haven’t looked at your calculations in detail yet, but I hope to later this week.
Mathematically, I’ve always struggled to come up with an optimisation criterion. At one stage, I was thinking to switch up a treasure when it no longer gives you 3 levels. While using the Treasure drop-rates and Gold costs should be the correct starting point, I’m not sure how to balance them.
For the record, here’s what I’ve used, depending on my Hoard Level goal (Treasures are numbered from 1=Common to 6=Mythic – “4x 3” means 5 level 3 treasures, 4 times):
L100: 10x 2, 10x 3, 33334, 2x 4 (1.3m Gold)
L100: 3x 2, 4x 3, 33334, 2x 4, 44445, 55566 (546k Gold)
L155: 34444, 7x 4 (L88), 44445, 4x 5 (L138), 6 (1.13m Gold)
L200: 34444, 7x 4 (L88), 44445, 7x 5 (L162), 55566, 2x 6 (1.97m Gold)
(Levels in brackets are so I know when to switch.)
All of these should yield Quality 10 automatically. The first one is expensive, but uses only low-level Treasures, so is a good option for your first few Delves, before you have more Chaos Shards than you need.
Level 155 is a great target, because it gives you an extra point of Magic from rounding if your Kingdom is level 14. Same goes for L177 and L199.
I’m aiming to get all my Hoards to Level 200 (199 would do, but … you know … round numbers…), and it should be the minimum for any Delve you’re working to max without Potions. But it does take a long time to get that many Legendary and Mythic Treasures.
One thing you’ll note with all my patterns is that I use smaller treasures early on to avoid unwanted overflow at the end. I also have patterns for boosting L100, L144 and L155 to higher tiers, if anyone’s interested.