Inductive charging can be done at > 90 % efficiency with realistic geometries for EV charging: https://ieeexplore.ieee.org/abstract/document/6953459/ . Foreign object detection is one of the main challenges in this field, the actual power conversion is a solved problem.
For those without IEEE access, they get a measured efficiency of 96.5 % with a coil spacing of 110 mm. The receiver coil is 240 x 240 mm in size.
Impressive if they can take that from the lab to a commercial system that operates at similar efficiency. It's not a verified real world working system though, just because you can get that efficiency in a lab doesn't mean you can mass produce it that efficient economically. Would be interesting to see independent test of that bmw-charger for example.
Even if they succeed with that, it's still 5% losses, compare that to a connector. Would you want to make your car 5% less fuel efficient just so you didn't have to plug in the cord? Unless they manage to get the losses negligible this isn't a good idea imho.
It's a classic example of an over engineered solution to a non existent problem that in reality just makes things worse.
Note that this includes the power conversion electronics. The best conductive (wired) charger DC/DC converters are presently around 97 % efficient, so the difference is small, especially considering that most chargers are between 92 and 94 % efficient, including typical 3.6 kW onboard chargers.
They only mean the electronics involved in the wireless transfer, so compared to a normal connector you have to add the 5% losses from their
laboratory prototype to the other charging component losses.
Once you have the power in the car, it's the same whether it was transferred wireless or not. After that you will need the same type of DC/DC converter to charge the batteries.
Ideally you should compare it with a system where you replace the wireless transfer module with a normal connector. It's the extra losses in the wireless transfer compared to a connector that is the problem here.
That there exist crappy chargers just shows that if these wireless chargers become common, in practice they will perform way worse than that.