The argument that this is a security issue isn't very well fleshed out either. As far as I can tell, it boils down to his opinion that this makes YAML harder to read and thus less secure. But, the reality is we have to copy & paste config today and that's a process I've seen fail when a change needs to be made and isn't properly carried forward to all locations. I suppose I could argue that's a security concern as well.

Half the argument against supporting YAML anchors appears to boil down some level of tool breakage. While you can rely on simplifying assumptions, you take a risk that your software breaks when that assumption is invalidated. I don't think that's a reason to stop evolving software.

I've never seen a project use any of the tools the author listed, but I have seen duplicated config. That's not to say the tools have no value, but rather I don't want to be artificially restricted to better support tools I don't use. I'll grant that the inability to merge keys isn't ideal but, I'll take what I can get.