The proposed alternative is identifying your domains organically while onboarding them into your architecture. This approach might be quicker but could risk complex operations of moving or remodeling when things start to break.

I have seen the consequences of this approach, and I would not recommend it. This may require so much reiteration, which would only lead to a big failure. So please stay away as much you can.

Most companies adopt Data mesh without fully understanding Data mesh concept/architecture. 75% of the companies are doing rework after building a large data swamp (lake). Organizations start by only hiring a few data engineers and One senior guy to lead them. They are not aware that this is not a data or analytical project. This is cultural change. So there are some prerequisites to start with. I always recommend doing an architecture assessment before beginning any significant initiatives within the organization. And I would always involve different architecture people, including Business, information, application, and integration architects of the organization. But sadly, this is not the case in 90% of the organization. That would be an excellent point to mention in your article.