I might have gotten a bit ahead of myself because I am not quite at 4w/kg yet but now ever more so determinded to get there!
The question asked was ‘what does it take to reach 4w/kg’ and I say the answer first and foremost is not ‘volume’ or ‘genetics’ but several years of consistent training! And maybe reasonable age 
For pro level fitness of >> 5 W/kg it will be a different story, though, most likely genetics>consistency>volume>age.
But I actually find the chart by Couzen to show a relativly weak effect of CTL on w/kg. As it was stated before you’d need to on average double your Fitness/CTL from 50 to 100 to raise w/kg from 3.5 to 4.0. A humongous increase!
And since the RMSE of the data set is ~0.58, a random atthlete is only moderately less likely to to sit at 4w/kg with 50 CTL vs. another athlete to sit at only 3.5w/kg despite a CTL of 100.
This also means that I am currently not an outlier, nor will I be if I manage to reach 324W at 81kg, because Couzen would predict me to hit 3.66w/kg with 44CTL, so only a deviation of 0.34 (well below 0.5)
The question is, if the amount of CTL is not a phenomenal predictor of w/kg - what is? I’d argue it’s most important to have a CTL at all in the first place. Only then other factors like volume, training type and quality, nutrition, recovery etc. slowly come into play. Probably with great individual variation.
To constantly have a CTL, it’s also very important to prevent excessive fatigue and thus burn out.
I might have missed it but his chart seems to also not account for age, gender, weight and prior training experience (e.g. newbie vs. seasoned athlete).
That 55yo ripping off legs on your local group ride almost certainly did not get there by doing 12hrs/week for a year or two. It’s been consistency over years and years, probably decades, that got him or her there.
So far, even though progress has slowed down quite notably I am still eking out new FTP PBs in my fourth year of TR on similar volume.
I believe it cannot be overstated that consistency is key and that a sustainable training volume is more important than a high training volume.