Terrible back-of-the-envelope math disregarding any kind of loss or inefficiency, and assuming a magical DC->DC power system which would convert a car's 14.4V alternator source to something suitable to charge a Tesla without losing anything:
A "normal" modern car's alternator can put out ~100 amps at 14.4VDC. That's a maximum of about 1.5kW, which will charge a Tesla at a rate of about 2 miles of driving per hour of charging, again assuming a magical efficiency.
The reality is that the car's own systems take at least a few hundred watts and alternators aren't built to run at maximum load all the time and are likely to fail after a while running near capacity. Plus, the Tesla doesn't, to the best of my knowledge, contain any sort of DC->DC hardware capable of bringing 14.4V up to the necessary charge voltage, and even if it did losses would be at least on the order of tenths.
So, even with a good, running car and a theoretical alternator, you'd be waiting a very long time next to the running car (and burning a lot of gas) before your Tesla would go anywhere.
Obviously it would be possible to build, for example, a tow truck with much greater charging capacity and as time goes on we'll probably see at least a few of those in real life. But the fact remains that gasoline provides an excellent energy density in a convenient liquid form and that for certain edge cases, its conveniences will be hard to beat.