Hi everyone, got some news regarding this story of KNS80 and curve track more or less accurate to what may have been planned,
here are his last result and finding on this technical case
Ep. 2
"I spent several evenings examining this problem, and even implemented the spherical geometry solution for my RNAV instrument. I quickly discovered that the spherical Earth was not the culprit, and it is not even worth shipping in my code, because the maximum difference I could produce at 200 nm from the VOR was around 5ft lateral error in waypoint placement. After several more days of testing, I have determined that this is the result of incorrect declared VOR magnetic declination. If you are to fly a route created with the KNS-80 that follows actual VOR’s so that they can be monitored with the RMI, the same angular error is observed. Unfortunately, a modern GPS confuses the matter, by using the declared declination from its database to “adjust” apparent radials when in OBS mode. This means that the GPS will give you the correct desired track to the waypoint in GPS CDI mode, but an apparently incorrect desired track in OBS mode that follows the named VOR radial! This is not a MSFS error, as I have confirmed the behavior in the GTN 750 Trainer.
If you’re thinking that this upends most of what we’re taught as pilots about flying VOR radials, following Victor Airways, and triangulation, you would be absolutely correct. A centered CDI in no wind conditions does not necessarily mean that you should fly the selected course to navigate directly to the VOR. This is only true when the VOR’s declination is accurate to the local magnetic variation. In the case of the Corpus Christi (CRP) VOR used in your example, the magnetic variation in the area has drifted over six degrees from the VOR’s declination since its installation. For some real world reading on this subject, the best explanation I could find was this report from the Civil Air Patrol: https://savannah.cap.gov/media/cms/VORRadialsvsMagVariation_0CFCC08B02E70.pdf . While there is no solution to this fundamental VOR problem, as it is completely accurate to real world behavior, there is lots of related discussion on this MSFS forum thread with regards to LittleNavMap: A VOR was off by 5 degrees. List of problematic VORs?
In summary: The RNAV behavior is accurate to the real world, and matches the indications from a simultaneously monitored VOR. The apparent angular error is introduced by VOR stations that do not broadcast with updated magnetic variation. In the real world, this is most likely disregarded as a wind correction angle, but absolutely affects real aircraft. This is not an error in the operation of Black Square’s RNAV systems in MSFS, and the courses should be flown as they would be to any VOR in the real world. See above for more reading. For VOR’s with inaccurate magnetic declinations, the actual magnetic bearing to the station is the flown course that produces a centered CDI in zero wind conditions, regardless of the OBS setting."
I m so glad to hear from devs that they are still working assiduously on this product in order to keep perfecting it 
as well I m impressed of the attention given to users feedback !
Thanks for taking time to explain and keep us informed about your progress.