I'd go along with Alexander's comment that working in SI units makes life a lot easier. However, assuming you have a good reason for sticking to US units ...
The torque you've calculated is in foot pounds. The easy way to see this is that the 33,000 conversion factor you've used converts horsepower to foot pounds min$^{-1}$, and the rpm is in units of min$^{-1}$. The min$^{-1}$ on the top and bottom of the fraction cancel leaving the units as foot pounds.
In the second equation you've put in the distance as inches, which makes life harder than it need be. If you take the wheel radius to be one foot rather than 12 inches you get the force equal to 682.77 pounds.
For the last step you need to be aware that the "pound" is being used as a unit of force here i.e. it's the force exerted by an object weighing one pound in Earth's gravity. In units of feet per second the acceleration due to gravity is about 32.18 feel/sec$^2$, so the acceleration of an object weighing 2712 pounds will be:
$$a = \frac{682.77}{2712} \times 32.18 = 8.09$$
and that's in units of feet/sec$^2$.
Maybe I'm just used to SI units, but I repeated the calculation using SI units and got the same result a lot quicker! Apart from anything else it makes the distinction between mass and force a lot clearer so you wouldn't have forgotten you need to multiply by the acceleration due to gravity to get the acceleration.
You're quite correct that if power is constant then torque is inversely proportional to rpm. However for most engines the power is roughly proportional to engine speed over a reasonable range, so the torque is approximately constant. If you look at the torque curves for family cars you'll see the torque is roughly constant in the middle of the engine speed range but tails off at high engine speeds. Very highly tuned engines have a torque that peask at higher engine speeds because they're tuned to develop a lot of power at high engine speeds.