In general, I found it a chore to maintain Go-based systems.
There's a lot to like about Go. But it doesn't seem pragmatic for the types of applications I see people using it for.
For example, the entire error thing is absurd. Anders Hejlsberg got this right many years ago: 9 out of 10 errors are "handled" by a central error handler (log + "sorry, try again"). You can do this with error values, but at best it's verbose. Maybe error values makes sense for system programming, I could certainly believe it makes more sense.
But, really, the main problem is the type system. I'm a huge fan of dynamic languages. I think they're so much more productive. This is especially true when your system is interacting with external data - like user input and a database. I did C# and Java for years. These interactions are, at best, painful with static languages. But they're downright infuriating in Go (poor reflection, no generics). I wrote a somewhat popular library to help deal with map[string]interface{} nonsense (2), but it's not enough.
When it comes to web/tcp stuff, I think OpenResty for any middleware-ish system and Elixir for everything else are far more productive and enjoyable ecosystems. We've been doing Elixir for about a year now, and I've never seen such a consensus on productivity and quality (everything else always had some issue: Ruby is slow, Node is single threaded, Java is verbose).
There's certainly more space than just that. Go might still be a good choice for CLI tools, and both OpenResty and Elixir have a distinct memory/threading model that precludes some use cases - but those are in the minority.
(1) - Nothing against Erlang model, but I do think it's more restrictive.