I think all of your points here are worthwhile.
Fissile material has vastly higher energy density. There is really no problem with creating nuclear fallout in space. There are designs for spacecraft which gain thrust by launching what are basically nuclear warheads out the back. The entire back end of the ship is an enormous radiation shield and shock absorber. The materials and technology to build such a ship have been around since the 70s. The trouble is getting it to space w/o detonating a dozen nuclear bombs in the atmosphere.
The deadly to humans part is complicated. It kinda reminds me of this XKCD What If?
Most manned rockets need to remain under ~3.5 g's if the humans are expected to be able to reach switches and controls (with considerable effort). During extreme conditions, especially a launch abort sequence, the manned capsule can reach over 20 g's for a short burst. This is pushing the limits, as 25 g's is the standard accepted threshold of permanent damage. There are a few recorded examples of humans surviving much greater accelerations.
@ bold: Hehe. No. You forgot about relativistic mass. The closer you are to the speed of light, the more massive you are. Meaning that it would take progressively more and more force to simply accelerate you the tiniest bit.
There is not enough energy in the universe to accelerate any massive object to the speed of light. Nor is there enough to slow any massive object even the tiniest fraction if it is already moving at the speed of light. Such an object would have infinite momentum.
Your estimate for 10% c is probably fine. Relativistic effects don't really show up until ~20 - 30% c. At 10% c, you'd expect your non-relativistic calculation to be off by ~0.5%.
Just to double check your calcs:
Spoiler:Time to accelerate to 0.1 c at constant acceleration of 1 g:
Using c = 3(10)^8 m/s and g = 9.8 m/s^2
delta_t = 0.1 c / g = 3,061,224 s ~= 35.4 days
Distance covered during that time:
Using 4.37 ly as the distance to Proxima Centauri
x_final = 1/2 * (9.8 m/s^2) * (3,061,224 s)^2 = 4.58(10)^13 m ~= 0.005 ly
0.005 ly / 4.37 ly = 0.0011
After accelerating to 0.1 c at a constant 1 g, you would be 0.11% of the way to Proxima Centauri.
(This is the non-relativistic approximation.)