I can see a reason to abandon BLE. Bluetooth is generally
a bit problematic maybe due to is complexity. but “Bluetooth low energy” is perfect for what it is (a super slow short range wireless communication implementation).
There's a user experience benefit of the semantic separation between BT and Wifi; you move around and change SSID all the time with a portable device like a phone, which is for connecting to "other" things. Your BT connections "travel with you" as they are part of a personal area network.
There are weird boundary conditions (connecting to your friend's BT speaker, using wifi to cast to a chromecast or AppleTV) for which neither is perfect; frankly I think those wifi cases are more like the BT use case.
Just because the two use the same BT spectrum doesn't mean they should be unified.
Their first revision was 20-23 byte packets. The first thing anyone looks for when doing BLE is “how to send larger data”. BLE 4.2 and packet length extension helped this but it’s still pretty rough because you aren’t promised any specific MTU size by spec, just hope you can the packet size you need on connection.
I have a BLE product in the wild, but the amount of work I did to get around packet limitations was probably not worth it all said and done.
BLE is good for event based data transfer of flags and small data, but you outgrow that pretty damn quickly.
I have a new interest in BLE on account of myself recently becoming bluetooth enabled.