I'm concerned that the debate about flexibility doesn't really get at anything important, and that there may be in some poster's minds a conflation of flexible with good. So that noted, thinking about what I mean by flexibility.
I mean, I really tried to get on top of exactly that thing above, spending rather a long paragraph talking about how the rules one uses to play and the style one pursues are genuinely separate and need to be understood as such.
To give a (probably horrible) analogy, let's compare two carpenter's toolboxes.
Box A contains a jigsaw, a hammer, a crescent wrench, a grout removal tool, a plane and a 1" chisel. Six tools.
Box B contains a single phillips screwdriver. One tool.
On this occasion, I need to attach one piece of wood to another using philips-head woodscrews. Box B is right for my purpose. Box A contains nothing that will do this job as effectively.
If we are speaking about individual, highly specific situations, yes, exactly this result comes out. But is that relevant for TTRPGs?
One of the big draws of the medium is that it can do an enormous variety of things. Even games I have referred to upthread as relatively narrow—such as Monsterhearts and Masks—are open to a staggering variety of stories. If they weren't, the (sub)genres on which they're based (
bildungsroman by way of "supernatural romance" and "superhero comics" respectively) wouldn't be
absolutely stuffed with the things. That freedom means one needs to be prepared for a variety of situations. "I genuinely need to do only one, singular and specific thing, which requires a specific tool," is exactly the opposite of considering a
variety of situations.
Or, to appropriate your analogy:
both of these toolboxes are bad. The first one is bad because it lacks a very commonly-used tool (screwdriver), the latter is bad because it literally only has one single tool, even if that tool is quite commonly useful for a typical suburban family. One would, in fact, expect that both of these tool sets, if sold as such, would be considered inadequate. Every toolset needs, as a matter of business, to aim for the fewest tools they can justify (because that saves money) while still being diverse—one might even say flexible—enough to meet customers' needs.
A grout removal tool is pretty specific, as are some of the others. Not like a saw, hammer, crescent-wrench, some screwdrivers (preferably flathead, Phillips-head, and at least one or two more), some kind of sharp implement (e.g. utility knife), etc. These are things likely to be useful in a variety of common contexts--and thus things likely to be useful in a tool set. (A
jigsaw specifically might be too much, but a compact hacksaw would make sense and be easy to store.) Such a toolbox would be "flexible," in the sense of handling a lot of different circumstances.
A knife is a flexible tool. It can do an awful lot of things. It is not a perfect tool for every possible circumstance, which may mean creating a new tool is warranted. A knife may be useful for doing such a thing.
That said, an issue with the analogy is that each tool really does only one physical thing: hitting stuff, twisting fasteners, cutting through materials, twisting other kinds of fasteners, etc., and the end-user cannot meaningfully change this without going out and buying more tools. In theory you
can use your tools to make better or more specific tools, that's how new tools came to be in the first place, but it's an
enormous amount of work. Not so for TTRPG stuff. Even for 3e, (in)famous for trying to have prodigious discrete rules, many rules cover multiple kinds of situations. It becomes quite possible to build a
reasonably comprehensive set of rules-tools, and to make a tool that sweetly and simply creates
new tools. (Or...less sweetly and simply.)
To take your question to an extreme, is an RPG with no rules the most flexible?
I would say no—because it doesn't do anything at all. To again borrow your toolbox analogy, that's like saying that box C which contains
nothing whatsoever is more flexible than either of the boxes provided, which is ridiculous.
Regarding armchair designers, isn't flexibility contemplated solely in view of modifications to the design? That is to say, no test of flexibility can be made where there is no flexing being done? Whether a system asks or does not ask users to be designers has no bearing on flexibility.
Surely not. Several people in this very thread have talked about re-applying the existing rules, without change, to situations other than the ones intended. E.g. Fate's Aspects have been brought up as flexible. I myself have cited DW and 4e as containing flexible structures, not because you can rewrite them, but because an individual tool or set of tools is capable of covering a plethora of situations. "Defy Danger" is the most commonly-used move in Dungeon World specifically
because it is supremely flexible. Skill Challenges are flexible, able to apply to a huge swathe of relevant situations. 13A Montages. Etc.
