The introduction date is not what matters - it is the date of discontinuation. Controllers often remain on the market for many years before being dropped. It is more difficult to justify supporting a controller than disappeared from the market ten or more years ago than a controller whose protocol continues to be implemented in products available today. HUI and Mackie Control have become industry standards, with controllers using those protocols still in production today.
I have been part of a team having to make decisions on technical debt, albeit in networking rather than audio software. There are only so many developer hours available each week. Old code that nobody has touched for a while adds complexity and can impede desired changes to internal interfaces.
I would be very surprised if Steinberg’s underlying direction of travel is anything other than full MIDI 2.0 support in a forthcoming release of Cubase and Nuendo; I suspect the new MIDI Remote was designed with future MIDI 2.0 support in mind. MIDI Remote will be the actively developed and supported route to add controller support from here on.
For now, the existing generic remote remains, but has been tagged legacy; maybe its feature set will eventually be folded into MIDI Remote (perhaps even with an automated update of existing generic remotes to MIDI Remote) and the legacy code dropped.
My point here was that the situation has changed since support for the Houston was reinstated. The push on the forums to reinstate Houston support came immediately after the release of version 11 (I think it was actually made in the Nuendo 11 forum). At that point, MIDI Remote was not available, as it was dropped from version 11 shortly before release so that it could receive further attention for version 12.
Your request for the legacy SAC 2.2 support to be brought back comes three maintenance releases after Cubase 12 has been released. MIDI Remote is now available, which allows for a clean re-implementation of SAC 2.2 support. It might be much harder to reinstate the legacy SAC 2.2 support now, as the internal interfaces might have diverged from those that the legacy code requires.
I can understand Steinberg’s position being “if the users cared enough about this, someone would have said something 18 months ago”.
I’m not unsympathetic to your request at all. I understand your desire to keep using your still functioning controller as you always did. On the other hand, I could understand the developers saying “we dropped this legacy code deliberately as we continue to tidy Cubase’s internals, nobody complained for 18 months, we’re not bringing it back especially now that self-help is possible via MIDI Remote”.
Version 12 has brought what is, for some, some painful losses. Whilst Mystic, Prologue, Spector and Loopmash had received no attention for years, they were important to some who miss them now that they have been dropped rather than updated for Apple Silicon and Steinberg Licensing support. Steinberg seems to have tried for years not to drop any features, but there comes a point when decisions have to be made about unmaintained code that is seen as non-core. On the one hand you have users who do not want any breaking changes, on the other there are those who criticise Cubase for failing to modernise as quickly as they hoped. Maybe Steinberg has the correct balance after all.
Others will be in a better place to answer than I, as I don’t currently own a Mackie like controller so am only watching on whilst others work on this. However, I am aware of active development on some Icon controllers and on the Behringer X-Touch series. There is a long forum thread about the work on the Mackie mode of the Icon Platform M+, which should be a good starting point for other Mackie-based controllers. There is a Presonus Faderport 8 script (for the non-Mackie mode of that controller) in the Steinberg midiremote-userscripts GitHub repository, which would also be worth reviewing. @Martin.Jirsak has been working on a Mackie C4 script. The Steinberg repository is open to contributions and should become an increasingly valuable resource over time. To be clear, we are not talking here about scripts generated in the Cubase/Nuendo GUI, but scripts programmed in JavaScript.
A good Mackie script should provide much of the groundwork for the SAC - how far it gets you depends on how Mackie-like the SAC’s native protocol is. From what you say, its scribble strips are an enhancement over the Mackie protocol, but the same is true of the Behringer X-Touch. If the SAC MIDI Implementation document turns out to be out of date, looking over MIDI dumps should be enough to figure out what is wrong. An incomplete implementation document that contains some errors puts you in a better position than having to reverse-engineer the protocol from scratch.
It would have helped if Steinberg had shipped a reference implementation of a generic Mackie controller in JavaScript, as that would have been an excellent starting point for implementing many Mackie-like controllers - but there is only so much developer time available.