There are a lot of assumptions hidden in "the market always works" statements like that. In this case, there's the assumption that the line you're quoting represents a business model that's cheaper than "buy the cheapest fridge possible and replace it with another cheap-as-possible fridge when it breaks," especially when you factor in that you no longer need a repair crew on the payroll (because you don't repair things).

Here's another, more complicated problem: even if the math proves that, say, over a long enough horizon, you'll make more money with the longer-lived, repairable version of the product, that isn't necessarily the best way to maximize returns for shareholders/owners. If that concern overrides all others, then you may well be better off maximizing profits short-term, through any (hopefully legal!) means possible regardless of the effect on the long-term viability of the business, and investing them in other markets with higher returns.

Not necessarily, depending on all kinds of market factors, e.g. like stickyness of the transaction: how lazy would consumers be about changing out a fridge provider with poor service, or locking in the consumers with time boxed commitments. The slower and more inconvenient the transaction, the more the provider can wedge in rentier profits, and the poorer the market will respond.

Sure, but then that kind of company starts to look more like a fridge repair business than a "RaaS".