I certainly agree with you that there is tension between powerful interesting magic items and keeping character ability in the forefront, but I think you've missed the point that obviates the conundrum: what decoupling magic item power (whether large or small) from character ability would actually enable if done well. This isn't about deciding whether D&D evolves to favor one of trivial or character-defining items, it is about evolving D&D to the point where each table picks whatever balance they want and the system as a whole continues to work. You can even do both in the same game in the sense that the dominant factor varies from character to character, or at different times.
All this is dependent on strong guidelines that are reasonably robust, and sufficiently detailed to enable the new DM to run the game successfully. I think that is challenging (and definitely requires thinking about the math) but not impossible. If that is achieved we have many decisive qualitative gains: Low magic works. High magic Monty Haul works. The rich inexperienced kid traveling with the freaking Vow of Poverty monk actually works. Matching risks to rewards works (and for most people would be the usual case) but regardless of the details of that relationship the system doesn't care. All the system should care about is what items and abilities the PCs have right now and provide tools so the DM can challenge that party. That's it. In my opinion the single greatest conceit of D&D is that the party (almost) always faces challenges in keeping with their present capabilities, and the game is aided if it can help us achieve (or knowingly violate) that conceit with the fewest possible additional assumptions.
So let the rules of acquisition, and player expectations thereof, be governed by the story, the setting, and the table. For two editions they have been baked into the system itself, but these particular cookies have turned stale.