They didn't have an APU (weight), so towing to runway would need an external spin up too. It would probably have no air or hydraulics until it got there.
Engines and intakes were designed to be supersonic efficient, which meant they were horribly inefficient at subsonic and taxi speeds. SR71, XB70 and similar big supersonics of that generation had similar, but worse, issues. Being military it didn't matter - they used to just in-air refuel the SR71 right after takeoff!
The SR-71 also had the problem that its fuel tanks wouldn't seal when the plane was cold, since naturally the whole thing was designed for sustained cruises with skin temperatures of hundreds of degrees. As a result, it leaked fuel like crazy before takeoff.
There is https://en.wikipedia.org/wiki/TaxiBot. In 2011 Lufthansa also did some tests with an electric taxi system in the main gear of an A320 but I couldn't find anything more recent on that.
More vehicles on the runway equals more chances for FOD on the runway - which ironically caused the Air France accident and loss of one Concorde and all hands on board.
Interestingly, the opening paragraph of the wikipedia article contradicts this:
> The TaxiBot eliminates the use of airplane engines during taxi-in and until immediately prior to take-off during taxi-out, significantly reducing aircraft fuel usage and the risk of foreign object damage.[1]
This "Taxibot" seems to be designed mostly on the tarmac where engines are idling to keep the APU happy and reach operational temperature, a place where FOD is often caused by humans, not machines - it's not meant for operations to the runway, where engines usually are at full capacity and humans are absent.
Most likely, in this different scenario, there actually is a FOD Advantage.
It needed to be refueled right after takeoff, before it had gone high altitude. In part because it never took off with a full fuel load (to limit strain on the machine and risks in case of takeoff flameout), in part because the fuel lines were literally leaky on the ground (to allow for thermal expansion, and because the fuel was used as coolant and fuel lines ran throughout the machine) and in part because the J58 had been designed specifically for high-speed high-altitude operation, they got more efficient with both speed an altitude as they switched from turbojet to ramjet regime. Sub-sonic SR-71 burned fuel like nobody's business to say nothing of takeoff itself.
> So basically the travellers would cause more pollution in the 10 minutes of taxiing than in maybe 10 hours of driving a car...
That's an exaggeration. 100 passengers, so 20kg/passenger, i.e. about 20l, which is less than 200km worth of fuel. Maybe 2 hours' worth.
And a car has pretty bad fuel consumption going in and out of the driveway too. You want to compare the fuel consumption for the flight as a whole really, since the alternative is presumably driving a transatlantic-equivalent distance.
Wow. Just wow. These engines are obviously far from their designed power band when taxiing, but this is incredibly inefficient.
So basically the travellers would cause more pollution in the 10 minutes of taxiing than in maybe 10 hours of driving a car...