I'd use a variable blade, fixed throttle turbine - ideally without a condenser
for space and weight reasons (although just 400C may be too wastefull in terms
of water consumption - higher temperatures would be much better). Safety will
be a problem. We have to keep exhaust temperatures well below 100C at any point
touchable from the outside - so planing the expansion path would be tricky but
probably not impossible within the boudaries of a standard footprint. It would
have to be a serial hybrid, of course. Running such a slowly responding source
directly would be ok for a locomotive that can just dump whatever heat it
doesn't need through a chiney and doesn't have to stop at a red light every
couple of seconds - but not with a car.
We built prototype hybrids with gas turbines more than a decade or so ago - so
principally, yes, it would work. And it wouldn't even have to be significantly
more expensive than a Chevy Volt (but we'd need a larger battery).
Despite all that, I don't see this technology in any car soon (provided it
exists). In my daydreams I rather see it as a large scale, extremely cheap heat
source that could make the energetic disadvantages of hydrogen disappear - and
making a hydrogen car/plane/cement-factory/blast-furness/whatever is probably
much easier to do than trying to cope with the disadvantages of a slow/low
temperture heatsource directly.
