Alright. So, I will give you my opinion for what it is worth.
The entire purpose of rigging is to give you the ease of animation with logical options to solve production problems while animating.
Rigging is broken up into two categories. 1) The process of designing the underlying mechanics and 2) the process of designing animator controls.
The background behind auto-rigging is simply and nothing more than the process of creating a module that can be easily duplicated and replicated across multiple characters - in production. The way this works is to set up a way to have a definition rig (the one that lets you adopt to the scale and proportions as well as add or delete options for a specific model) that can be edited and then create scripts that 1) create the underlying controls and 2) Creates the controls for the animator.
A good auto rig is the result of years of research and work. Built upon decades of previous R&D, standard practices and known issues and solutions. Rigging of course is a constant and evolving field. When new technology and tools within Blender (or any other app) become available an auto rig can be updated or replaced to include this technology.
The purpose of new technology is - in the end - ease of animation and solving more problems in production which results in improved results while animating and faster approaches to save time.
There is absolutely no situation where a rig is too complex or “overkill” unless 1) The underlying mech is flawed or 2) not enough thought was put into the animator controls - including giving one the option to turn things on and off as needed. As well as options to adopt the set up process to a particular character. For example not going through the tedious process of rigging and setting up deforms for a fleshy character when you rigging animating a robot. This is an option, not a requirement.
A rig properly designed (1 and 2) results in the best animation experience. And the best animation experience is speed, flexibility and quality. Probably in that order of importance.
It has already been predetermined by decades of those who have gone before that the effort spent on rigging (research design and implementation) pays off 100 fold in animation speed, flexibility and quality.
So if you are going to embark on animating characters it is important to understand this ratio of effort. Something like 100 to one. (just to pull a figure out of my arse But it is probably more like 1000 to 1). Animation is incredibly tedious and time consuming and is by far the most time consuming (labor-wise) of all of the production processes in 3D animation.
And the decisions you make about rigging should reflect this ratio. Otherwise you will pay for it in increased and almost deadly animation time. For real. You want to kill a project? Go lax on rigging. That will do it.
This of course is aside from the various time saving techniques of animation you can learn.
Now here is another ratio to consider.
The time it would take you to research understand, experiment with and create rigging at the level of something like Blendrig (or rigify) that gives you a proper ratio of advantage in animating time is probably again 100 (or more) to 1 of the time it would take you to research and study and understand and existing system.
So time invested in understanding the fundamentals of rigging to the degree you can then study rigify or blendrig and learn to use either system is but a mere fraction of the time it would take you to create your own.
And going back to a comment earlier. And if you were thrown off by my comment about drivers then you have a long way to go to create something that is worth the effort and will give you the ease of animation and not a headache trying to use a tool not properly designed.
My opinion comes from 20+ years in 3D. A lot of time studying rigging as well as rigging my own characters. And time spent research my own approaches. Not to mention experience with auto-rigging systems on many platforms.
I am by no means an expert on rigging. But I understand enough to know where the effort should go.
So in my opinion unless you want to become a full time professional rigger, and create some cool stuff, stick to animating.
And further I think there is a reason they went with BlendRig and are not using Rigify - to my knowledge.