It feels to me like GitLab might be shooting themselves in the foot by spreading their resources too thin, too quickly.

They have the very grand ambition to take over the entire devops toolchain from source control to deployment, and some parts of the toolchain can certainly benefit from the tight vertical integration that GitLab provides, and when everything just works the product does end up providing a very streamlined user experience.

However, this also means that their monolithic toolchain has to be able to out-compete best-of-class specialized tools in every part of the chain, and every part of the chain where they fail to do so could end up as a potential dealbreaker for a large subset of their user base.

I've used GitLab.com for a few projects, and in general I've found the various aspects of their product sorely lacking in terms of overall polish, performance, and capability compared to excellent specialized products like GitHub for VCS, CircleCI for CI/CD, and Pivotal Tracker for project tracking/planning, etc.

Also, the fact that they're trying to take over the entire toolchain means that just about everybody else in the devops space sees them as a direct competitor, so until they grow to become an 80-pound gorrila that nobody can ignore in at least 1 part of the toolchain, very few other companies building devops tooling will willingly integrate with their product, which makes them even less attractive for users with specific/more advanced needs that their product alone can't meet.

With all that said, I really appreciate their dedication to open-source and am personally rooting for their success, but I can't help but shake the feeling that they're fighting a losing battle. DevOps tooling is simply too diverse, complex, and opinionated a field for any 1 company to dominate from end-to-end. Focusing resources to become the best tool for the job in 1 area and using that leverage to foster an open-ecosystem around that product seems like a much more viable long term strategy (though that could certainly prove to be difficult if you're trying to become the best tool for VCS, due to how entrenched GitHub is in terms of developer mindshare).