I think the base assumption is that the adapter will be "dumb" adapter and all the logic built into the car itself. Unlike CHAdeMO, they are not going from a smaller amount of pins in the car socket to more pins on the connector (in which they need electronics in the adapter to simulate the expected signals in the extra pins). SAE actually had a slide about this in a presentation.
One possibility is that future Tesla cars could have the signalling protocol built into the car (as many of us believed it was going to originally), needing only a simple mechanical adapter for the plug shape - which could probably handle the locking issue by some kind of hinged flap such that inserting the plug+adapter into the car locks it. Would need a hardware upgrade on existing cars however.
For future Tesla vehicles, yes. Current ones can't generate the required signalling (so far as we know - and I think we do know now), and generating it in the adapter will need a fair bit of power - much more than can be leached from the pilot pin.
All interesting and valid points. My take on the situation:
I can see how you could provide a mechanical linkage inside the adapter that was acted upon by the cars locking solenoid and effectively locked the CCS pin too. This would solve the locking issues, but also negate the need for electrical power.
Easy for a small passive adapter, quite a tricky mechanical design job for one in the form-factor of the CHAdeMO adapter, but probably solvable as you say.
However, it doesn't negate the need for power unless the signalling has all been implemented inside the car - the chipsets for the carrier-based signalling probably need a watt or so.
I think CCSs spec has provision for CANbus too. In fact this is how the Supercharger protocol works as I understand it (the guys at EVTV did a video showing how the trace logs they took tapping into the cars wiring during a Supercharger session). There's an initial PWM handshake, then the EVSE and the car between them agree to go digital, and use CANbus. This is in the standard, but AFAIK only Tesla have implemented it.
Not unless you've found a standard that I haven't. The standards and Tesla agree with the PWM signalling on the pilot pin to say 'go digital', but Tesla interpret that as 'go CANbus' while the standards say 'go to Greenphy carrier and exchange XML-formatted messages'.
I'd go further and say Tesla are VERY close to being within standards for Type 2 DC mid charging. The biggest contention is they are drawing a higher rating than the physical standard the pins were rated for. Of course this is a 1hr continuous rating, and SC are not at full power throughout a charging session, owning both the car and the charging equipment they have provided a level of engineering they are comfortable with from a product liability POV. A public charge station manufacturer, where any car could turn up and simply draw max current for extended periods does not have the same luxury and is constrained to the letter of the standard. (Is this why Elon walked away from the standards committee? )
Any charger manufacturer can instrument the temperature inside the connector (indeed, they are required to do so by at least the SAE version of the standard; I can't remember if it's explicit in the IEC version). So they could if wanted implement a higher initial current that ramps down with connector temperature. Doing so doesn't impact cars that aren't expecting this scheme, as they won't ask for that much current in the first place.
However I think the simplest thing for the EU market would be simple installing a CCS port on the car. A redesign of the rear tail light would probably give enough physical space to fit one. I suspect this will be mandated sooner rather than later for WVTA rules into EU anyway. IOW it won't be the CCS chargers that drive this, but the type approval rules for new vehicles. (2017 will see the end of Type-1 ports for all new models, and I suspect all the manufacturers faced with moving away from Type-1 to Type-2 may very well just go the whole hog and opt for CCS.)
It will be interesting to see how this plays out. A more rational outcome would be to end up on the Tesla connector (ie. standard type2) and the carrier-based signalling, but it's easy to see political factors that stop us getting there.
BMW drivers seem to be having a lot of trouble with the Ecotricity CCS units where they have to hold the connector 'just so' to get it to work. That doesn't sound like a problem to be solved by a software upgrade, though I suppose it's possible that DBT and/or BMW have got the connector geometry slightly wrong.