Whelp, I finished. 100% conquest of both Mondstadt and Liyue (version 1.1.1). Full disclosure; I have been running a modified version of the game, mostly so I can fix any incidental bugs I discover, so I can ignore those issues. While most of the bugs are already known (training and the wharf's buildings) I did notice that there are a number of hidden locations on the map. I assume that's intentional? The issue is that there's one of these hidden locations that's also completely inaccessible, but still counts against your occupation of Liyue - Mt. Qingce.
Now, for my actual observations;
The new difficulty changes are.... actually OK? Without the enemies scalling from your own training, it's mostly fine. I did have my Liyue garrison force get overwhelmed occasionally, Elite Milleliths are a special kind of hell to deal with. Even had my garrison units get overwhelmed a few times because they got jumped by 2+ of them plus some elite xia support. And I think that's a good thing. The main issue I had with the new system is that it also seems to buff "boss" enemies. And that's probably a mistake. You shouldn't really need to bring a squad of 4 lawachurls and 4 megachurls to kill Morax. But due to how that fight works, you need to be able to blitz down his shield and kill the dragon before they can re-deploy the shield. And since the threat buffing seems to increase shield health (or defense applies to the shield, so it's taking less damage), it makes the fight very unballanced. At a minimunm, boss monsters shouldn't receive any threat buffs. Personally, I'd say that any heroine and one-off boss creatures shouldn't receive any threat bonueses. Because otherwise those fights are going to be really hard to balance.
The new colonization system is a wonderful addition to the game. It generally takes a lot less resources to get started than the old system, and I find it adds a number of interesting choices to the game - namely what buildings and colonies to prioritize. You'll get every building eventually, but the order is important. The only real complaint I have is that you can't see all the building nodes initially. Even if you can't build them yet, knowing that you'll need X number of a specific slave is very important information. Particularly if you've also got a training option that uses the same kind of slave. Don't want to spend all your elite knights on training before you actually build the building that gives you them, after all.
The farm is an interesting addition. Lot of choices in terms of what you can build. But one of the things I found myself doing was re-building things a lot. I initially built some Dark Wedges to generate abyssal power, but later phased them out for ley line sprouts. I built some basic farms, but later upgraded them to abyssal farms (and replaced some with ley line sprouts). Basically, what you want from your farm in the late game is different enough from the early game that you need to do a fair amount of rebuilding. And being able to upgrade farms to abyssal farms directly, or recover spent wedges, branches, and the other tree items,would be helpful. Particularly if you are removing the elemental trees from your farm. It also feels like the late-game farm is a bit uninteresting. Since there's so few sources of abyssal power, and you need so much of it for buildings and units, It feels like the best late game farm is just 30 ley line sprouts. But I feel that that's more due to how other systems currently work, rather than how the farm itself works.
As for potions and food? I find them more annoying than anything else. The main issue is that they are both only temporary modifiers. Which means you either need to remember when they turn off, or check back every week to make sure they are still running. I'd rather that these were implemented as a toggle. You'd still only be able to have one active at a time, but you wouldn't need to do any micro-management to keep the bonus active. For potions this would be relatively simple, but for food it'd be a bit more complicated. Food currently use "nector" for the more complex resources, which is produced by planting Whopperflowers. Which isn't exactly conductive to having a weekly upkeep. But - there is something already in the game that does and would thematically fit; colony item production. A lot of the expensive colony buildings produce a food related item; the windmill produces wheat, the winery produces wine. It'd make a lot of since if those resources where what was required to make certain food items, instead of just nector.
Finally - the teapot. It's a trap. The bonus elemental power is helpful in the mid game, particularly if you use Lumine to boost all of your elemental resonances. However, you can get a surprising amount of power from just corrupting heroines, sacrificing units, and beating certain units. To the point where I'm not sure how valuble getting a consistant trickle of ~2 elemental power per week actually is. And while you can use it to generate Abyss Power... with the vast majority of heroiness having between 40-50 power that's only going to generate +2 per week per person. You'd need 50 weeks to repay your initial investment.... and that's not counting how much power you'd be able to generate by breeding those characters and sacrificing the results in the same amount of time. Which is why I feel it's a bit of a trap - in the late game your colonies end up producing a lot of advanced units, to the point where you don't really need to keep breeding characters for a supply of the basic elemental hili/mita/sama-churl units. And since thier randomness doesn't care about your elemental power, elemental power for breeding drops of in utlility later on. And once you get buildings that can produce lawachurls? You don't really need to keep heroiness uncorrupted. At which point you only really care about nycto unit production... and your nycto elemental power is reduced by your building expenses. Including the teapot.
Which brings me to something I'm a bit concerned about - there's more things in this update that cost abyssal power to use and build, but there's not as many good sources of abyssal power. The only real source is from destroying statues, at 200 a pop. But we no longer generate power from building wedges and branches, instead needing to plant them. And while you can get a lot of weekly abyssal power from that... it takes a lot of resources to build up to the 50 per week level we could reach in the previous version, and at a greatly increased risk level. Not to mention that nycto units have a built-in negative feedback loop not present in other units. With nycto unit upkeep reducing nycto elemental power, meaning that producing more units reduces the quality of future units.... unless you sacrifice them. And while rebalancing nycto units to be spawnable at lower elemental power levels, that doesn't change the fundamental issue that you need 2000+ more abyssal power than your upkeep to reach ~100 nycto elemental power, something that's relatively easy to achieve for every other element. And this is because the sources of abyssal power roughly keep pace with your abyssal upkeep. Destroy more statues? Get more heroines you need to capture. Plant more wedges/branches? Well, now you need to start using samachurls on the frontlines. And if you do manage to get a nycto mitachurl? Well, that's going to cost more upkeep to store and deploy. I just feel like this system means that you really have to dedicate your entire build and approach to generating as much Abyssal Power as possible, all so you can get enough in the bank to start producing quality units around week 175. And just changing the threshold values for generating nycto units doesn't really address the underlying problem of how nycto elemental power is calculated.
Plus, there's also the price of wedges and branches. Those items are less useful now than they were in the past, and they are required for more things. But branches still cost the same to produce, and wedges are twice as expensive now. So their cost could possibly use a re-balancing. Just throwing out numbers here; maybe 50 for the wedge, 250 for the branch? You'll need a rather large stockpile of wedges to keep your frontline units reinforced, until you can get colonies established nearby. and colonies need a massive supply of branches to finish. I'd also suggest tracking item progress seperatly, so item progress can be retained when switching production. It made sense previously to wipe progresss. You could only make one item at a time, and branches where just better wedges. So you'd basically build wedges until you reached over 100 production per week, then switch to branches. But now that they are needed for diffrent buildings and do completely diffrent things, you end up switching between the two items a lot. And it's a bit of a pain to lose all progress when doing so.