I think that when people see SQL Server they see a very visible and huge cost. I'm finding that even those who you have to squint a bit when they class themselves as millenials are reluctant to adopt the disciplines needed to put together a decent DB design.
These people are quick to jump to the NOSQL options and a palette of technologies to "solve" their problems but many of those problems are of their own making.
In my experience a lot of the costs in alternative technologies are hidden. Few think of the integration costs or that technology 'x' does 90% of what is required so now you need technology 'y' as well. Each component introduces a new thing to learn, a new thing that has to be supported in some way and a new thing to integrate into the whole. It isn't a trivial exercise to get all the components up, running, stable, playing nicely with each other. And then a component needs to be upgraded and that can be painful even with DevOps approaches.
There are some awesome pieces of software out there but there is a part of me that wonders how many of them exist because a problem has been created? How many solutions are actually coping strategies for choices that looked after the interests of a few in isolation rather than the bigger picture.
There is a lot that I wouldn't use SQL Server for but it's like rich mans shoes. A pair last a lifetime whereas a poor man spends far more because his don't last.