If not, is there any schedule to improve FilletEdge to this level?
Not automatially with FilletEdge because both fillets run off the two surfaces they would follow. But yes, the shape can be created using FilletSrf, trimming, and joining.
I hope it will be on the future list
plasticity and siemens nx both use parasolid kernel.
Can you send those models and filleted results as step files?
Yes, I do not have access to the system at the moment
Iām also wondering how that performs on non rational surfaces.
The Blendcorner command is also very useful in NX
This is too good to be true. I refuse to believe it, even if itās a totally existing tool. 
lol
@Mehraneftekari thanks for the samples. The quality of the fill surfaces in the first samples is excellent for the most part (in the half indented cylinder there are a few tangency issues)
As for the second part: impressive result, but it looks weird. That one could be done in Rhino in two steps though with a different topology, by first making the concave fillets and then the variable fillet with FilletEdge.
Can you post a photo of it? Because there is no problem in NX
Thanks
If you explode those models in Rhino and turn on the control points of the resulting patches, you will figure out that they are extremely dense. For example, one of the surfaces in āNX_Fillet2.stpā has 600 control points, wile another one has 560. Makes it very difficult to edit manually (if necessary) and increases the file size.
Yes, I tried
looks beautiful!
āblend cornerā is what Iād call ātrue filletsā , but some ppl probably like the huge nob before itās blended and call that ātrue filletā 
Not to mention the surface edge cusps before the blendā¦
I wish it could fillet it, as sometimes there are extra splits when objects are created with Rhinoās tools, which make filleting difficult.
I think @pascal is the expert on Rhino filleting here in tech support.