We're just going to disagree on this and leave it. You're starting to ignore parts of my statements and only reply to the parts you like. That gets tiring.
I do my best to respond to everything, but have limited time. Sometimes, my only opportunity to reply is on my phone without my glasses on, and I think my posts are so disjointed and riddled with typos, it's hard to understand what I mean.
I think this is going to click for you, so, my interest isn't to ignore parts of what you say.
The article I reference above to Buka supports the things I'm trying to say. It's not that good training is irrelevant or has no impact. It's that application of training is critical, and in particular how the training is applied, the timeliness of the application relative to the training (i.e., how soon did you apply the training after the training was received?) and the relevance of the training to the task.
Here's another way to look at it. The training isn't there to replace application. It's there to facilitate the transition from not doing something to doing it. Let's say there's a bridge. On one side is incompetence and on the other is competence. You can put a blind fold on, and ride across the bridge on a unicycle. This is like self teaching... people learn to do things all the time on their own. You may not succeed... certainly, you've stacked the odds. But people get to the other side all the time without any help from training. Just figure it out. Training allows you to walk across the bridge. You'll get to the other side faster and without all the drama. It won't be as exciting, but you've increased your odds for success. But the bridge is still just the bridge. Competence is on the other side. Going back and forth on the bridge never gets you to the NEXT bridge. You'll never even see it.
Everything overlaps. So, while skill development is linear and predictable, skillsets aren't, particularly when you start getting into complimentary skills sets. It's more like the following (if I can make this make sense): The point I'm trying to illustrate is that performance in skill 1 (whatever that might be) enables one to begin developing skills 2 and 3, which then enable one to train in skill 4. And while that is happening, the person is continuing to apply skill one, and is moving beyond performance into expertise.
This same things can be applied at a skillset level, on a more macro scale. Larger context, but it's the same thing. Cops gain skills in one area of self defense. MMA guys gain skills in another. Each could approach self defense training from a different angle, because the skills they're actually acquiring are complimentary.
EDIT: Just want to add that the reverse is true, too. Sometimes, to get to the point where you can begin to train something, you will need to be performing or an expert in several other things... So, the pyramid below expands based on dependency of skills or skillsets, but also contracts similarly, based on dependency.
[TABLE="class: brtb_item_table"][TBODY][TR][TD]Skill 4[/TD]
[TD]Skill 2[/TD]
[TD] Skill 1[/TD]
[TD]Skill 3[/TD]
[TD]Skill 5[/TD]
[TD]Skill 6[/TD][/TR]
[TR][TD][/TD]
[TD][/TD]
[TD]Training[/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD][/TR]
[TR][TD][/TD]
[TD]Training[/TD]
[TD]Performance[/TD]
[TD]Training[/TD]
[TD][/TD]
[TD][/TD][/TR]
[TR][TD]Training [/TD]
[TD]Performance[/TD]
[TD]Expertise[/TD]
[TD]Performance[/TD]
[TD]Training[/TD]
[TD][/TD][/TR]
[TR][TD]Performance[/TD]
[TD]Expertise[/TD]
[TD]Innovation[/TD]
[TD]Expertise[/TD]
[TD]Performance[/TD]
[TD]Training[/TD][/TR]
[TR][TD]Expertise[/TD]
[TD]Innovation[/TD]
[TD][/TD]
[TD]Innovation[/TD]
[TD]Expertise[/TD]
[TD]Performance[/TD][/TR]
[TR][TD]Innovation[/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD]Innovation[/TD]
[TD]Expertise[/TD][/TR]
[TR][TD][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD]Innovation[/TD][/TR]
[TR][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][/TR]
[TR][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][/TR]
[TR][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][TD] [/TD][/TR][/TBODY][/TABLE]