Note that paralleling transistors can be a good idea or a bad idea. With bipolar technology devices (BJT, IGBT), there's a positive tempco effect which causes devices which are conducting more current to become less conductive, thus reducing their power dissipation and cutting the amount they conduct. The load naturally balances. With field-effect devices, the opposite is true- if one device starts to conduct a little more, it heats up more, which causes it to conduct more, etc., until you end up with one device handling way more current than it should, and that one burns out.
Note, also, that the driver IC supplied with the Adafruit MotorShield
is a BJT based device, so you can stack them to increase current handling.