I think with anything like this there's always going to be some risk that somethings been overlooked - hence why we test things.
However, I think in this case, whilst the technology/ability has been there for some time, Specialized obviously decided not to implement it because they were no doubt worried that people might break/brick things. Also - with anything like this, once you open the software up to be able to update the firmware remotely using a third party device, it's easier for someone to hack the code and then theoretically start creating and downloading their own versions of the motor firmware - which will no doubt lead to broken motors.
One big difference to Shimano and Giants implementation is revealed in SRC's comment :
"
We'll keep the exact details of update content and tests for those willing to participate, but it goes without saying that a) we've tested already and are satisfied that the process is robust and b) that any BETA user can roll-back firmware during the BETA test period."
This would suggest that because of the way the firmware is applied/installed, what it updates, it doesn't update/corrupt the equivalent of the "BIOS" (basic input output system) - the most simplistic bit of software/firmware running on the bike, which in this case also gives access to bluetooth. So if you do an update and it doesn't work for some reason - you haven't bricked the bike. You've just installed some faulty software on the motor or the install went wrong for some reason. Fortunately, you'll still have access to the bike via your phone/tablet so you'll be able to roll back the firmware update or update a different version. It does look like the roll-back option may be removed once this is out of beta, by then presumably they'll be confident they have a system which works properly and there aren't any ways for things to go wrong .. in theory