I don't have a huge amount of data for running, all I know for me (so far) is I can plot CTL / 5k performance reasonably neatly. So far, more is more. I suspect, there's a cut off with this. In terms of you max out your aerobic capacity and you can probably get by with less CTL, for the same performance and add in hills, speed work. I'm not there yet, I don't think.
My guess, as would be with cyclists on time crunched (mainly being top amateur/ sub elite time trialists in the UK) that you probably won't reach the point where this is a huge factor. I'm talking 9 hours a week or less. 8 hours is the point in which I've got in my head I might have reached diminished aerobic returns and have to try something else to add in. But time will tell.
A guy way smarter than me on the old TT forums, did his own study of a number of people's training and how the acquired CTL matches with performance. The general jist was, the difference in FTP at a given CTL for each individual, was probably +-10% spread, if you consider what is the absolute worst training schedule you could do (my z2 for weeks and weeks and hours and hours on end) versus the best bang for buck, being sweet spot. Basic example , CTL of 60. A sensible, generic training plan might generate a FTP of 300. The worst training you could do, just aimlessly riding, but generating the same CTL, might get you can FTP of 270 (note it would take you way longer in time to also reach the same CTL) . A really well thought out, sweetspot block of 6 months might scrape out 330w ( in much less time riding). All for the same CTL of 60. That's the extreme end. For me, I was a much narrower range , where X CTL basically equalled Y power from a baseline of about 160-70w untrained and a CTL of 0.
The guy I talked to a lot about this was a swimming coach, he found it also applied to swimmers he'd known and triathlon specialists. To me, something along these lines makes perfect sense for running, albeit with some small running specific factors etc, mainly the ability to absorb pretty big punishment on the body. But in a nutshell, it's roughly what I've stuck to, followed the plan, the numbers and I'm still seeing the improvements I'm happy with and would expect to see.
Incidentally , again, this would need bigger studies, there are two people who messaged me direct on Strava (I think one has posted in this thread) who have adapted this to do marathon pace work , so right at the bottom of the sub threshold range and they have racked up huge CTL numbers in 6 + months and without even really thinking about it. One set a 5k PB and one set a 10k PB. Sadly, neither had accurate historical data from years past, but the main takeaway was from talking to them, that almost certainly in terms of overall load, they were both just doing much more than they ever had, but just in a controlled manner/ racking up the training load. Much like I was shocked when I suddenly hit a power pb after a big crash and just riding the turbo aimlessly to keep myself busy.
I stand by that nearly a year on, it's probably likely to make most people improve, but as we have seen with some cases in the Strava group, it's very easy to get it wrong and come unstuck like other training modes, if you go too hard. Ultimately, most of us here are hampered by our aerobic capacity. Training like this is pushing it up from below, even at what seems like pretty easy running levels.