I think it comes down to the fact that batteries are big and heavy, so they end up being pretty integral to the structure of the car. A strong quick release /engage mechanism for that sounds rather difficult to do. Especially where the car can't flex when full of people/stuff during the swap.

Tesla tried battery swap and abandoned it. If even early adopter Tesla customers won't accept this idea, it's very unlikely the mainstream will.

I suspect it would have to be imposed on the market through regulation and standardization. Otherwise you'll see endless quibbling on the swappable module specs, or 30 different types which would make it impractical to stock a replacement station.

of course it needs to be done right

The lost route IMO was modular batteries. Say your Model S has a bank replacable modules instead of a single monolith. These can be pulled and swapped with pre-charged cells in minutes. This also allows older packs to be pulled from circulation, removing the fear of aging batteries causing depreciation on the vehicle.

Why would you ever not want to swap the whole battery? They've demonstrated that a whole battery swap can be done in minutes. And the whole battery discharges/ages at the same rate, so there'd be no point in only swapping some of the modules.

It's also quite unsafe to run lithium ion batteries in series with different levels of wear / charge state, the old ones can easily be pushed too hard and fail (spectacularly).

I figure that if you have different-size cars with different battery capacities (and shapes of their battery cavities), then "Car X needs 10 battery modules, car Y needs 13" is simpler (in one sense) to manage than "I have car X, do you have any of the batteries made to fit in it? No, you only have Car-Y-shaped batteries? Whoops."

The angle I was considering was size and weight. If you can keep the modules light enough to carry by hand, it means a lot less equipment for each swap facility.