I suppose the problem with trackers and other demoscene-related stuff is that they are hobby projects. People use their spare time to solve a problem, but that probably starts out as a simple project, and ends up becoming much larger than anticipated. So there was probably not a lot of planning and documentation involved. It just grew organically.
I suppose another thing that could play a role is that often these people are still very young and inexperienced. Back in those days, the average demoscener was probably 15-20 years old, so they would not have completed their education yet, let alone that they had any work experience in software development. They were just talented, self-taught individuals, who were just getting started in their careers.

Having said that, the demoscene did indeed produce quite good documentation and specifications for many tracker formats, and on programming sound hardware etc (especially for hardware where the SDK was not free or commonly available).

I personally don't release a lot of my demoscene source code, because it probably generates more questions than it answers.

Scali! But have you given a try to set online some demoscene code fragment which can be something to start? Are there even those sites (nowadays the search engines get pretty confused if trying to seek something specific and out of the ordinary).

At first (after such research and social engineering too) ... I hope there is some hope. Both software AND hardware wise.

There would be time to have in the forum also the section for literature (not only (mostly unobtainable) books for sale, but rather complete manuals for download ), for a programs and device drivers development kits for DOS, etc. These manuals should be complete downloadable set, not just webpages, as the www is pretty good to vanish, changes itself to be incompatible, no printerfriendly.
The good examples which to take as example would be the old Asphyxia ones, and somewhat unique Turbo Pascal newsletter where all was together for offline browsing and all printable too.

About trackers remakers...I have seen already all those clones to die one after another and the authors take its source to the grave. THe history has shown the development taking its (collapsing) course about such programs "rapid but undocumented" development.

What if it is just about the hardware price and availability, not just about the corporations decisions which set the stop for new program development?
It is not just "becouse it is old and slow" issue, why otherwise there is still old hardware which is getting very good exposure, there are always coming new software, projects (intel8051, PIC16F84 ..) and also for this slow slow Arduino (compared to GHz machines), tons of software projects for these things, yet they are not in a 1GHz region, nor PC compatible! And many books.
Okay, perhaps not fair comparison, but hey, DOS old soundcards are far from win10, just as ZX Spectrum is from PC, and yet there ARE those who make software for it still now!
Still, how come that someone comes to idea of these 8..16Mhz Arduino things! In a time with gigahertz CPUs...?? It requires proprietary software and has a bag of tricks of its own. It is not a PC, not ZX-spectrum. Then why not the same thing with retro soundcards too and many other below 50Mhz things... 😉
So, what makes them "tick"? (no pun intended).

Is the trick that no-one can as big corporation can derail compatibility and dictate or enforce "upgrades" or "updates"??! While WINtel in its combination just removes PCI bus, the same cant be done with old MCU chips as they are still in production and the new ones which are faster are backward compatible? Also these are widespread, and cheap to obtain, every bit is known AND welldocumented.

Also these are widespread and there are new ones too. So, I imagine someone still making the programs for ZX80 and for C64 too.

Trouble is with PC the old hardware is not up to the enthusiasts to take control, they are not found in abundance as PIC16F84. Instead of that we see inflated prices on the old hardware). And retro kits... just making it for sale, well, but those are not sold in masses.
The floppy drive emulator is now from China 14EUR with shipping. Can someone beat that price in EU or US by making these kits?
Hey! Shipping itself costs already so much in Europe, thus feeding 7 middlemen and increase the prices. This can also be the reason to obtain new computer which is relatively cheaper than feeding the retro hobby and dealing with the software.

I think the one obstacle can be here that why these retro things have not got so much popular is that some retro things are relatively expensive becouse of these seven middlemen with hairy hands.
It is due to the shipping fees in EU and US compared to China. The detail costs. While we are not yet counting the labour. But everything which comes from china can be sent with very low price and 10 chinese salesmen competing with each-other will bring down prices until it is so low that even details are more worth than that, (thus getting the card in order to get the details for next prototype experimenting).
So, the escape hatch would be here to let the chinese multiplicate in zillions after all is ready.
Just before that it needs to be modular enough.This way perhaps it can make something (the next thing we need is in abundance those 486s).