These days, it sort of depends who you ask and how you measure them. Some people think the HD650 is neutral while others think the HD600 are a neutral reference. You can sort of see why if we look at this smoothed graph from Sonarworks. For reference, Sonarworks buys like half a dozen pairs of a headphone and uses their average to determine the frequency response they have to correct for.
But for sure, the HD600 is still probably considered to this day as a reference headphone.
RE quality control of the X2. I'm not really in the know except for the fact they started using glue to attach the earpads rather than use friction fitting. Apparently they stopped doing this in response to the internet wide outrage? I dunno, I'm someone who is typically suspicious of any audio impression. Has anyone objectively proven that there is an actual significant sonic difference between old and new X2s? I'm not saying that people are lying but people have made some pretty insane claims that have been more or less proven to be false.
Like when Sennheiser started to provide graphs show no 6khz peak, Head-Fi went nuts claiming newer HD800 are less bright than older HD800 samples. Yet third party measurements showed that there was basically no change in frequency response that couldn't be attributed to listening position or earpad condition...basically if you hated the HD800, you should still hate the HD800 regardless what serial number they were. If people started liking the HD800 based on the serial number, then that should tell you how powerful bias can be when it comes to audio. The glue bullshit + manufacturer is probably enough for some people to think like this.