Nicholas’s approach prohibits local detail which is the whole point of adaptive tesselation, isn’t it? It also won’t work well with strong creases (only at very high polycounts) and will make it difficult to sculpt areas that are close together (limbs…). It will again require the artist to progressively add resolution or face bad performance and result in uniformly tesselated meshes that have too many polygons in some ares and too few in others. Personally, I think UC looked vastly superior. It’s a pity Farsthary won’t work on it anymore.
I don’t think he’s said he won’t work on UC anymore.
Seeing BMesh merged to trunk is surreal - I remember meeting Briggs (original author) at BConf 2007, and me thinking it would be 1 year out 
As a long time OSS user I’ve learned not to hold my breath, but I’m glad I wasn’t told back then that it would take 5 years!
Thank you to everyone involved making BMesh a reality, much respect! I look forward to the future with this new foundation.
Is this official or are you guessing?
Well no, but he works for Pilgway now, on a product that directly competes with what UC does. It’s a conflict of interest. For all we know he might be just saying that to hold off development by someone else (that’s not actually my opinion, just a possibility).
Already adjusted myself to the new bmesh methodology, and I am loving it. I seriously missed ngons in Blender (coming from other apps), and now it’s finally here. The new knife tool, dissolving, the bevel - modeling heaven. Especially for more mechanical objects for me - though even for characters I really can see the benefit and the ease for less destructive modeling methods!
Thank you devs!
Lack of ngons has a big reason why one would not use blender, glad it finally made it, that will certainly bring in a lot of new crowd who like me love ngons. It will also liberate us python developer when we make blender mesh generation addons, code wise ngons are far less easier to handle than trying compute all in quads and triangles especially when dodgy topology is involved ,
can you imagine blender with bsurfaces that have ngons? YEEEES
Now someone can tell Farsthary that his “waiting for Bmesh” unlimited clay integration can go ahead among other things.
Does Farsthary still works for pilgway(3dcoat)?
Bmesh is cr*p unless the bevel improves.
Yes and it has been months since I’ve seen something Blender related posted on his blog. C’est la vie - but is normal that he focus on his work on 3D coat as that brings food on the plate.
Well, it’s a good thing that bevel will be improving in the future, eh? 
knife, connect, dissolve, bridge, ngons(<-its amazing quick to move poles about when the are no triangles in your way)
Mmmh, this is like saying Blender as a whole is cr*p, because bevel was not even present in the 2.5x-2.6x series until now.
Don’t even know if this post is worth a serious answer, though.
As far as I remember, nicholas’s approach will have local detail. What you’ve seen so far is just an early demo.
And I don’t think other things you mentioned will be problematic if he does it right.
Looking very forward to it; preferably one without quads. Waiting for Bmesh has been the reason for the delays so I’m hoping it will follow fast.
that is exactly what I asked in the bsurfaces thread 
, Bsurface with ngons and the new boolean can revolutionasize the speed of mesh creation in blender. I also see an interest in improving metaballs , which can be rely on the same code as remesh modifier
how much work will need to go into bsurfaces ,boolean to get them to work with ngons?
well as far as bsurfaces the author said in his thread that he is already working into integrating Bsurfaces with Bmesh. Booleans are an included feature of blender so I guess they will be bmesh friendly if not already. So “soon” is a pretty good estimate from a first look.
This is exciting. I’ve been waiting for something like bmesh to be implemented in trunk for a long time so I can start transitioning my polygon modeling workflow to Blender from other software. And it opens up so many other methods that can be built on top of it. This is my first post and I bet you’ll get lots more new converts like me in the coming months!
I am getting triangles when using the knife to draw a square inside a quad face and I am also getting a hole. Dissolve vertices also creates holes. Its not a big problem though I have find ways to get around those problems, most of bmesh seems work okish for now and I am sure its a matter of time before its fully usable.