I had used puppy since 4.3.2 when it soon was replaced with Lucid, great improvement. I adopted the early puppy-studio by lowt3ch, and enjoyed all of these distros. However, I never found them to be an all out replacement for Windows, as the linux world used to be, until recently, somewhat difficult for a "typical user", it was just difficult to get specialized hardware working, among other hurdles that mainstream ms/apple software did not present.
That was until fosspup64_9.5.
After a couple of months of using it, I felt puppy had arrived, and I promptly erased windows from every machine I owned and haven't looked back.
It became clear though, after a couple years, where the achilles heel in the "puppy-concept" was for me, and that simply had to do with a puppy being fixed to a specific set of packages of a certain upstream release, and that the OS and it's software simply had too short a shelf life in this fast moving enviornment.
So when the First-rib distro KLV-airedale came out, I tried it, and it was an "aha" moment for me. It was completely different, maybe not a puppy as I had come to understand, but certainly a forward looking approach. So I jumped on early while still in the beta phase with the thought that as it progressed I would fully understand how it operates. And that has been the case. I now use F96 and KLV variants exclusively, though have booted many other forum distros, these are my daily drivers.
The KLV model has a few key ingredients going for it which I will attempt to clarify.
1) Conservative, dependable, and reasonably current rolling release model upstream. The Void ethos to me is very compatible with the puppy sensibility. Not bleeding edge, but offering flexibility as well as choices in things like systemd, xorg, wayland, pipewire, alsa/jack etc. The fact that Void is not cranking out a completely new base system every couple years is particularly compatible with puppy sensibilities. In other words the one Void rolling release has a longer shelf life with great customization potential and adequate backward compatibility.
2) The first rib install tools are seemless. wd_config provides dependable boot stanzs. The layered filesystem involving only two digit number prefixes makes custom additions of layers, either compressed or uncompressed, totally configurable by a power user. Multi-instance install script makes experimentation an easy reality using minimal disk space. The Psuedo-Full Install technique allows an average guy like me the ability to build a system without relying on mastering scripts and configurations, and it also allows one to run a system as if it were a full install while still having no need to partition a separate area of the drive where it resides. The two techniques combined create a layman's development architecture that has no equal so far, to the best of my knowledge.
3) @fredx181's backup/restoresys scripts included in the KLs allow one to take a Psuedo-Full-Install, customize it, build it with desired applications, then back up the "upper_changes" which are in reality the entire rootfs excluding the kernel and modules, and then squash that backup into a read-only compressed filesystem (a new build) all while still running in the system. Then using @wiak's multi-instance-install technique one just needs to link to that filesytem in the other instances and reboot, at that point running in a newly created system, read-only and compressed. In fact I do this about every 2-3 months. I run the compressed read-only multi-instances and leave the instance0 as a Psuedo-Full install and untouched. After experimenting in the other instances, I simply boot up the Psuedo-Full Install and make the tested changes there, back it up and resquash the back up. New daily driver created and ready to symlink to the other multi-instances.
These key ingredients make the KLs very powerful and compelling distros.
I feel they should be in the "ready-for-prime-time" space, where ever that is these days.