3GPP has been pushing 5G use cases for years now, trying to come up with a killer requirement than can only be satisified by 5G and a mobile packet core approach. Private networks are one example. There genuinely are cases where wifi doesnt cut it, eg in heavy industrial environments with lots of reflections, signal loss etc. Nokia do good business in this space for example.
Problem 1 is that getting access to spectrum is a pain. And if you are the kind of enterprise prepared to go cellular, you're also the kind of enterprise where your network absolutely needs to work all the time. So your spectrum access needs to be guaranteed, which is not something that you can easily get today (eg CBRS in the US). The alternative is that you get a little chunk from your local mobile operator, but dealing with an MNO is absolutely miserable for enterprises.
Problem 2 is that wifi is not standing still. Wifi 6E and 7 are picking up juicy lessons from the cellular world such as preemptive handover between access points as the client moves around. Wifi chipsets are dirt cheap compared to cellular ones so the cost of devices is lower, and sometimes good enough is better than best.
Problem 3 is that big enterprises like to have consistent BOMs across all operating entities. To my mind, the only bands that can be used on equipment all over the world with no meaningful roadblocks are those that run in the ISM bands, eg the wifi bands and the (?)850Mhz one used by industrial gear. So qualcomm are trying to bring 5G to a new spectrum band that isnt universal and where there is already a cheaper solution. Gonna be an uphill struggle.
5G is designed for private spectrum where interference is minimal. I wonder how well it handles interference from other devices on the ISM spectrum.
By the way, there are better WiFi APs. Ruckus’ APs have software configurable antennas that are rather amazing at improving SNR. It works by becoming a semi-directional antenna by choosing an appropriate antenna pattern. This technique is in addition to the 802.11 standard’s beam forming. I have had excellent results with them in my home after keeping the 2.4GHz channels widths at 20MHz and disabling the DFS 5GHz bands.
Problem 1 is that getting access to spectrum is a pain. And if you are the kind of enterprise prepared to go cellular, you're also the kind of enterprise where your network absolutely needs to work all the time. So your spectrum access needs to be guaranteed, which is not something that you can easily get today (eg CBRS in the US). The alternative is that you get a little chunk from your local mobile operator, but dealing with an MNO is absolutely miserable for enterprises.
Problem 2 is that wifi is not standing still. Wifi 6E and 7 are picking up juicy lessons from the cellular world such as preemptive handover between access points as the client moves around. Wifi chipsets are dirt cheap compared to cellular ones so the cost of devices is lower, and sometimes good enough is better than best.
Problem 3 is that big enterprises like to have consistent BOMs across all operating entities. To my mind, the only bands that can be used on equipment all over the world with no meaningful roadblocks are those that run in the ISM bands, eg the wifi bands and the (?)850Mhz one used by industrial gear. So qualcomm are trying to bring 5G to a new spectrum band that isnt universal and where there is already a cheaper solution. Gonna be an uphill struggle.