Schoolyard bullies are just that. Bullies in a school yard.
I forgot if I said before… credentials - BSc in computer science, German degree in business administration, professional experience working for a world class systems integrator, years of running money (of which a good part is judging companies’ or industries’ prospects with limited information).
Thank you @James for weighing in. 
Shooting from the hip - a well designed internal database with content ingested from all services would augment the experience a lot. I know, that takes work to pull off properly, but is also rather everyday stuff in software engineering. The integrations are there already.
Andre @Goodstuff, I agree market research is where it needs to start. The instrument of the buyer persona is very useful. One can look at current owners, past owners, ones on the fence and ones who are potentials but don’t know dCS. In the end there would be maybe 10-15 personas. It’s a bit complicated but so much better than just statistics. It also helps communicate ‘the customer’ across departments. A market research firm can do focus groups, interviews, questionnaires, etc. If I were an executive, I would want to sit in on some of the live things. The personas can be associated with market sizes to gauge attractiveness.
The personas can then be used for user story mapping. It’s important for that exercise to know what the user’s purpose is. For example, Mosaic works for ‘initiate playback’. It does not work for ‘interact with music collection’. My collection is not small and not large. Some 2,500 albums - 2,000 of them offline. Mosaic currently does not let me interact with that collection properly. There’s so much possible with databases, filters, metadata and views. Roon Focus is maybe 60% of what is doable with good design. Why shouldn’t I readily see my albums for a furious mood? I can look at the metal genre or the electronic genre but then I’m already looking at 300 albums. Yes, there’s probably no metadata anywhere readily available for that. But if the interface is done well, it’s a snap to drop albums into mood buckets once. Or the user story ‘discover music’ - it doesn’t have to be an algorithm. It can be content partnerships. I would actually much prefer that over machines choosing my taste for me. Or the user story ‘manage library’ - that’s a royal pain today. Buying the music, downloading it, tagging it, putting it into the collection. All fragmented and arcane (MP3Tag, Yate,…). There can even be a desktop component for library management. Subscription based to cover cost and a nice marketing tool for dCS. New people get to know the brand. The ones who have digital collections worth mentioning. Roon library management sucks by the way.
Roon is somehow positioned as consumption overload and the rush of media. Spot on a main trend in consumerism. Not dumb. Mosaic could be positioned for serious music lovers and collectors. If there are enough of them of course. This can be a real differentiator. Not just ‘a well done interface’.
Implementation comes down to people, process, cost. Always. Do I have the people? If not, can I realistically get them on board? Process is well established in software product management. It needs the person with the know how and skillset to implement and run it. That’s a dedicated job. Cost is an envelope determined by the usefulness of Mosaic in terms of long term risk management, customer loyalty, marketing potential.
Do I want to start developing a proper Mosaic the moment Roon goes belly up? That’s late. How many people do I lose, because Mosaic can be better? Marketing research can tell. How many new customers can I get with an upgraded Mosaic? Again, good marketing research will reveal that.
In the end, it’s probably a question of how much of a software company do I want to be. My personal view is, there’s great potential in exploring that route despite the implementation risks.
