Short answer: I don't know. Maybe the best way forward would be a sizing scheme that assumes most garments will be tailored to fit properly, similar to how when fitting for men's suits, the most important dimension is making sure it fits on the shoulders, because the rest can be tailored. For obvious reasons, this approach isn't really compatible with fast fashion.
Longer answer: I got so sick of inconsistent sizing that I learned how to draft and make my own clothes. Even when I learned pattern making methods, I found they were steeped in
standardization, which often required me to take a more ground up approach and build a custom pattern from scratch, for my body, rather than following a pattern.
What we need is the opposite of simplicity - I think the current problem has arisen from an excess drive towards standardization and our current sizing situation doesn't give us the words or the understanding of the variety of body shapes and compositions.
Bra sizes are a great example of this - the website bratabase.com is a community build bra database, and if you look at the photo galleries for a particular bra size, you'll see a huge variation in the boobs that fit those bras, both in shape and size. Ofc, some of these people may be wearing the incorrect bra size for them, but I don't think that's enough to explain the huge variation we see here (especially given that people contributing to a project like this are more likely to be aware of their correct bra size).
Having gone down the rabbit hole to understand the esoteric witchcraft of bra construction, it is utterly absurd to me that we only use two dimensions in describing bra sizes - the cup and the band size. Anyone who has gone questing for a well fitting bra will know that the process usually involves a fair bit of trying in different bras at or around the size you usually wear, and that's my point! There are entire dimensions of variation that we don't have the tools to adequately describe beyond rough advice like "Have you tried a plunge/full cup style? What about this brand?"
I'm going to tangent for a moment.
Back in the 1940s, the US Air Force's first fighter jets were experiencing a high rate of crashes that were neither attributable to pilot error nor mechanical failure, and they assumed that, because the cockpits were based on pilot measurements from the 1920s, that the dimensions of the average pilot had changed enough that the fixed cockpits no longer correctly fit.
A researcher took 10 of these human dimensions (e.g. chest circumference, height etc ) used in the design and measured 4000+ pilots against them.
- Zero of these pilots were within 15% of the average for all 10 measurements.
- And for any given 3 dimensions, fewer than 4% of pilots would qualify as average
This surprising result was super influential for more than just plane design. The air force began to "design to the edges" and required unprecedented levels of adjustability in designs going forwards, which manufacturers were very resistant to, but the air force were firm in their requirements and manufacturers reluctantly got with the times, because what are you going to do, lose the US fucking air force as a customer?
This set a precedent for adjustability in way more than just planes, once the manufacturers and designers learned how to overhaul their process to account for the added complexity of adjustability. (Summary source: https://www.gse.harvard.edu/ideas/ed-magazine/15/08/beyond-average)
Over 70 years ago, Lieutenant Gilbert S. Daniels' study showed that if you design a cockpit for the average pilot, you've designed a cockpit for no pilot. This was solvable through significant upheaval in the way that planes were designed and built. I see the problem of clothes sizing much the same as this old, solved one, except it's much more diffuse and therefore triggering the necessary upheaval of the not-working system is harder than it was for the Air Force.
Although I did recently discover something related to this that's very cool - https://freesewing.org/ , an open source software project that's building a library of parametric designs. I've not had a chance to try any patterns out yet, but it looks promising.
(I did say it was the long answer...)