A few years back I would've agreed with you 100%. I really do believe, though, that developers have worried about the plausibility of their mechanics so much that they've sacrificed good game design. Sometimes real life logic just doesn't make for good video games.
Its not so much about real life logic, then it is acutally making the skills fun to do. A minigame is one method, but this can often become tiresome depending on the game. Also playing Pipes to hack stuff is so random it might as well just roll dice. For example I found the pick-locing minigame in TES 4/5 was an improvement over Morrowind's stab-the-lock-until-result-of-RNG-is-above-some-numer mechanic. I likes the fact that they went to
some effort to make it seem like your sort-of picking a lock. However, I dont play these games for the hundreds of hours some people do, and can understand that the minigame could become annoying. It ties in with my main problem with RPG mechanics. I just don't like how heavily statistical they are. Sometimes it gets to the point where you can work out the probability of success/failure of a given action using simple maths. Its been like this for years though. Our 64bit 3ghz multi-core processors are still doing the same mundane logic that 8bit 4mhz cpu's were in the 80's.
I do, however agree that RL logic definetely won't always work in a game. There are plenty of boring things in RL that I dont want to do in a game. Hell, the reason I play games is because they are
not RL. As such game-design can't forget its a game.
Doing something a few times and having it matter is a rewarding experience. Having to button-mash 300 times just to be allowed to be better at it is pure grind.
The post was mostly an attempt to describe a method of making a skill fun and interesting. There are lots of other things that you could tie onto that, for example different types of materials with significantly varying and interesting differences (instead of material of +1 defence, slightly-better material of +2 defence). Of course mashing a button 300 times would be boring though this could be due to a boring mechanic, a skill/level/etc that advances far to slowly, or game design that focuses on the ends and not the means.
If the skill is boring to begin with, then of course someone would have no interest in increasing it. That person may prefer to do some other action instead. But IMO in this case the mecanic should not be necessary. Making armor can help you, but there can be other ways to aquire it, allowing you to advance. It should
also be fun. If the person doesnt particularly enjoy the mechanic but they need to do it for whatever reason, a reasonably original and fun mechanic shouldnt be painful to do, as long as it isn't required over and over.
RL actually has a pretty good idea, if you find activity A boring, go and do activity B. You don't advance in activity A, but given you dont
need to, nor do you find it entertaining to, thats fine. If you want (or in the sometimes-unavoidable situation need) a product from A, you can aquire it somehow through B (eg selling and money etc). Throw in some largely varying activities and I think it can provide fun, and a varied game experience.
Wheather any of this acually eliminates grind depends on the player though. But
hopefully at least it can reduce the grindy-feeling for most. What I described doesn't really eliminate having to do activities per se, then it does with trying to make doing them an interesting and original experience. What is interesting will be subjective, even if the activities are varied.