If SIMs keep getting smaller, how long before they disappear altogether?

A thought occurred to me whilst attending the recent SIMposium event in Berlin at the end of June. Given that there is a new standard​ being investigated/explored for an even smaller SIM card; how small can they get before they disappear altogether?

The new proposal, put forward by Apple for investigation by ETSI, the standards body, is for a SIM form factor even smaller than the existing 3FF micro SIM. The aim is to free up more space for other components within the handset body. However, on this basis, why not get rid of the SIM card and slot/connector altogether?

After all, with surface mount technology (SMT) being employed to embed SIMs within other equipment for M2M applications, the technology is there. Similarly, the ability to remotely provision and personalise SIMs has progressed quickly in the past two years. Numbers, operators even, can be allocated and changed after delivery/upon activiation - so why not do this with handsets?

I am summising and hypothesising of course. There is an existing ecosystem built around the current model which would take time to replace, plus MNOs would likely resist any changes, but it may not be too far from reality. And what would this mean in the bigger picture?

Potentially, smart card vendors would be forced to adapt and become software and services companies - something that some, such as Gemalto, are already moving towards. Meanwhile, the IC manufacturers could replace them on the hardware side of things, providing modules and SMT components directly to end equipment OEMs - something which they already do in some related applications, such as ID and payments. Could it be that market developments would result in the smart card manufacturers being replaced by their own component suppliers?