I really don't understand where you're getting that from. For economy cars up to ~130kph hub motors are a very feasible alternative.
There are hub motors+rims right now with which you could build a ~150hp car, that will weigh less or comparably to old-fashioned steel rims, including axles, rotors and superfluous suspension components (around 30-40kg). For example, the QS Motor 273 with a power of 28kw (per wheel) given appropriate cooling, at a mass of 43kg. This is for a mass market, aluminum motor/rim. I assure you that with magnesium wheels and glycol cooling you can get this down under 43kg. These motors have a wholesale price under 2000$, too, and have been run at 180v 600A for short periods of time.
Here is also an actual study with data about the feasibility of hub motors in four-wheel vehicles: https://www.proteanelectric.com/f/2018/04/protean-Services3....
Yes, hub motors will always have a disadvantage as far as unsprung mass as improvements affect both sides of the equation. However, hub motors have very significant handling advantages that you cannot feasibly get any other way, such as per-tire regenerative braking, a perfect fully dynamic differential, perfect torque vectoring, optimal weight distribution, allows far superior suspension design, and so on. As a result, hub motors are still a very serious contender, and as the gaps narrows and investment ramps up we might see very high performance hub motors become a reality.
Another technology that will have to be evaluated is fully active suspension, which eliminates almost all of the disadvantages of hub motors and keep all of their advantages. Indeed, fully active electromagnetic suspension is entering limited mass production right now.
But yes, in the 1990s hub motors could not have made for anything more than an economy car with a top speed around 130-140kph economically. Which is exactly what they were meant for.