The problem is that all else isn't held equal.
There are two ways to increase revenue per m^2. One is to intensify the construction. The other is to limit supply.
For numerous reasons, the latter approach ... seems to predominate. Rather than constructing denser residential, commercial, and (where appropriate) industrial space, the tendency is both to sprawl out, and to simply refuse to allow new construction. San Francisco is the pathalogical case, but hardly the only, and numerous other US growth areas (LA, Seattle, Austin) and elsewhere (Sydney, Vancouver, London) are to large degrees similar. There's new construction ... of large McMansions. But not of denser forms of housing.
By imposing an arbitrary constraint in one dimension (rent extraction per unit) but allowing intensification on another (density or height of construction), this should be addressable.
Keep in mind that most of SF is at best 2 storey construction. You don't have to airdrop Salesforce Towers all over the city, only allow construction to 4-5 storeys, to more than double effective density. Apply this more broadly over the SF Bay Area, and what's become a region-threatening blockage to housing availability (and incidental issues such as traffic congestion, hours-long commutes, access to schools and services, etc., etc.) would be lifted.
The fact that landlords (or banks, through mortgages) can extract arbitrarily high amounts of rent from a given area regardless of intensity of development, is the problem. The land value tax is one approach, and is probably a better one. I'm considering what the implications of a rents cap might be.
A usual objection to rent controls (increases on existing units) is that landlords are not incentivised to maintain or improve the quality of units. Whether that might also apply to a cap is a possible objection.