Best Of

@Alex_J - I really appreciate you taking the time to give me those thoughts, thank you (And good timing too, because I didn't do any work beyond adding those edge loops on the retop after I posted that!)
I completely agree with your closing statement, our priority right now is having a presentable slice of the game for our next deadline in a couple weeks to really sell the visual style and gameplay in a more complete state than our alpha submission, so I'm going to focus on polishing my texture and foliage work as much as possible and sweat stuff like proper optimisation later. I was definitely very hung up and stressed over doing the process the "right way", but you've given me some really useful insight from a different perspective I never thought to look through! I feel much more reassured with what I'm doing now.

As I'm writing this I've just baked the high poly of one of the trunks onto the ZRemeshed low poly that's been auto-UV'd in Substance and it looks more than good enough for now even if the technical aspects are probably sub-optimal. I'll update later today with the new polished texture work Thanks again! 

About the best way to unwrap the tree, I'll share a few thoughts: 

A good guideline is to not do any work unless you can clearly explain what problem that it solves. 
Another good rule is to always do things the simplest way that you can and only do it more complicated if there is a compelling reason to do so. 

When you are making a game the bottleneck is always going to be time. Peoples work cost money and every minute somebody is working, that is cutting away from potential profits. If you cannot explain clearly why the work you are doing is going to help in the long run, it is probably not worth doing (or at least not until you back up and figure out exactly why the work is important to do)

So with that in mind, what do we gain by manually retopo the tree? Is it going to give us a time savings? Will it increase quality in a way that makes a strong emotional impact for the player? If you cannot give a definitive answer to questions like that, then I would recommend to go for something faster and simpler. You could do an auto-retopo and an autowrap and then bake the hi to the new low. Whole thing can be ready to go in 20 minutes like that. 
This is quick and dirty and will not give you the absolute best result. However, it will probably give you a perfectly good result quickly and then you can continue moving more quickly to get other things done. 

Once the entire project is all ready and you've tested it and everybody is feeling good about what's there, then if there is time to perfect things with the art you can go hog wild with it. But chances are good that many things will change between now and then, so taking additional time to optimize art assets could end up being a waste. 

Dont fall for the common fallacy that if you do things in a more difficult or laborious way, that will mean it has more value. Working from assumptions - even if you mean well and are doing your very best - can make you a liability more than an asset. And doing things strategically - even if it means doing "sloppy" work at certain points in the project - does not mean you are lazy. It's a matter of prioritizing task so that you answer the most important questions in the project as quick as possible, minimize labor lost due to future changes, and push things that depends on others as far back as possible. 

note that there are plenty of good reasons to manually retopo and uv the tree, of course. It is just a question about when is the right time to do that. Getting to finish line a month early with a good looking game is better than showing up late with a halfway perfect game.

The Rig

Nice, a new challenge! Whenever I see a chest, I have to model
Probably will try to keep it simple (as always), maybe just geometry and diffuse, roughness, metallic maps.

Looking forward to see what people are making.

Edit: Started painting a texture atlas. Previewed on Sketchfab.


Edit2:
Continued with the texture atlas and tweaked some geometry. Used Blender and Sketchfab to preview. 
model
The original texture atlas is 1k.


Ended up hand painting some normal map, which makes for an interesting, exaggerated look imo. Maybe I'll experiment with that some more in the future.
- Texture atlas could be refined/optimized. 
- Unique AO on UV1 could be added.

kanga said:

Good info above. Anatomy is letting you down. You can practice using an app like Daz as reference. Free and the figures are customizable.

Concentrate on structure. Get a handle on generic figures first. You can turn off textures and rotate about the figure to see how forms connect.

Buzzkill, but this is not good reference. Do NOT use this for anatomy practice. @b_fair, Stick with actual 3D scans of real people (3d scan store, triple gangers, etc.). Don't use DAZ, unless you want to stifle your ability to read actual form & scale with accurate depth of how the computer sees the human body.