It's not an easy answer, and I don't pretend it is.
I'm using his sliders as a fantastic reference point and combining that with my personal feel from running through a number of seasons in a dev-and-sim solo franchise.
Extreme example which is easiest to illustrate my point: ignoring overall talent distribution and optimizing for the ideal number of 95-99 OVR and 90-94 OVR players at each position will accomplish exactly that. Take Kicker. At 100% slider settings I was able to take a 66 OVR rookie UDFA with 88 KAC and by midway through season 2, he was at 92 KAC. Spent 0 time developing the K/P positional group (and 0 people in a 32-man league will do this). So he's now 23 y/o and I'll have him in the mid-90s soon in the only attribute that matters. He's still only 69 OVR. Double his XP and I'd easily have him at ~94 KAC and still just 23 in year 2. Now you've inflated kicking accuracy a great deal based on data about how 32 CPUs develop their kickers and the assumption that OVR is the best measure for this.
In a solo franchise, OVR is probably fine. But we all know OVR is just a formula, it's one person's opinion of what attributes matter, and anyone who's ever played online knows for a kicker, KPW and KAC are what matter. AWR just doesn't really matter, but that's what will drive OVR and what the CPU is semi likely to develop.
So I used tdawg's data across all his sims (esp the early ones where he was listing 80-84 OVR guys and hadn't inflated sliders to 200-300% at nearly every position) and am making judgments based on that data + my own feeling & experience playing in & running leagues. It's not the best way to do it, but it's better than ignoring the truths about OVR and CPU/Human differences.
So I will be using a different set for our league, and despite my concerns about the methodology, I owe tdawg for his data, without it I'd be going 100% based on feel, which is akin to driving blind. At least now I can use his data as a reference and make adjustments from there.