Hi, I strongly agree, the SFF-ready is just to have a minimum bar that anything can pass
My take is this: compare TDP and size...
[MEDIA=reddit]sffpc/comments/12ne6d7[/MEDIA]
Based off this, my opinion is that cards should easily be > "200 Watts per liter or 200W/L". The R9 Nano is approaching 10 years old and easily did it, we should get better over time not worse?
So e.g. worst case (lol) I have a Meshlicious that could fit let's say a 4 slot 15cm wide 30cm long card
so 200"W/L" * (8cm*15cm*30cm/1000) = 200 * 3.6L = 720W GPU in order to satisfy this
but e.g. the 5090 FE has a great cooler design that's >300W/L so even at 575W (80% of this) it actually only takes 1.66L (45% of this), because it's >50% more "space-TDP efficient"
On the other hand the 5080 FE reuses the same dimensions cooler and so at 360W (50% of this) it still takes 1.66L (45% again), since it's only 10% better as expected
In an ideal world, we could also do some noise-temperature normalized testing, along with size
So my thinking is that if a card is larger (i.e. under 200W/L), that is "okay", but it MUST use that to be ("equivalently") quieter / cooler...or cheaper (the irony is that more premium materials would only be used on the small and large designs?)
And then if you want to get more crazy you could add in weight, since copper is heaver than aluminum but better otherwise? And people who want to travel with their PC might be willing to pay premium to get both "space-efficient" cooling and "weight-efficient" cooling.
It's depressing to me to see very tangible improvements in SFF CPU coolers but not in GPU coolers...Imagine if over 10 years your CPU cooler options became 2x larger / cooled 2x less... and people still called it a SFF cooler and recommended it (simply because it vaguely matches dimensions of popular cases)
