Pretty much all consumer APs seem to optimize for either a) cost-cutting or b) bleeding-edge standards & maximum theoretical performance. The reality is most consumers' experience would be better with even an order of magnitude worse peak speed if it fixed the ridiculous long tail latency & packet loss consumer APs in noisy rf environments have.
If you took hardware from like, Cisco or Aruba APs from two generations ago, licensed the OEM's automatic channel selection/rf tuning for that generation, made a little app which had you move around one of your APs for a week or so to let it gather spectrum data, & locked down most options so only your automated tuning could fuck w them, you'd already have hands down the best wifi a consumer could buy for any price. And if Cisco or Aruba or whoever did it themselves or your new company managed to get a favorable deal on the near-EOL hardware & its firmware, it wouldn't even be THAT much more expensive than the garbage we have today like Eero
Tbh the biggest reason this doesn't happen is that most users' ISPs are also absolute dogshit so even if you fix the mess on users' LAN they'll have tons of upstream issues, which makes it harder to sell consumers on spending more for better wifi hardware
Every layer of the stack normal people are subjected to is so so so much worse than they deserve. If the garbage we forcefeed them actually worked at all maybe users wouldn't have so much ingrained learned helplessness re: computers and they'd be able to actually benefit from FOSS! But obviously companies have all put profit first, and FOSS has put narrow, short-term ideological goals ahead of actually materially increasing user freedom, so here we are.
If we want to serve FOSS's (purported) overarching goal we have to stop playing stupid games over arcane IP licensing law and focus on making users' moment to moment interaction with technology not suck so maybe they'll have some interest in engaging with it more deeply
