In reply to henwardian:
The simple answer is the maximum possible reaction rate for the T-T reaction is about 25x lower than the D-T reaction. See this graph which shows the reaction rates for various types of fusion reaction versus the temperature of the system (measured in keV):
http://www.kayelaby.npl.co.uk/atomic_and_nuclear_physics/4_7/4_7_4b.html
This means that it is more difficult to achieve a self-sustaining T-T fusion against the energy losses from the plasma, especially when you take into account that each T-T fusion releases less energy than each D-T fusion (the ratio of energy released is about 60%).
But you also ask why the T-T reaction rate is slower than the D-T rate. This is a difficult question and one I couldn't find a perfect answer for. To answer it you need to delve into the theory for fusion reaction rates.
For the linked reaction rates graph, the curves can be approximated by the third formula (<sigma v> = 0.8052...) in the section on 'Fusion reaction rates' on this page:
http://www.kayelaby.npl.co.uk/atomic_and_nuclear_physics/4_7/4_7_4.html
The most important term in the equation to answer the question is A, and then to a lesser extent R. A is 56 times larger for the D-T reaction, hence it contributes a factor of 56 to the difference in D-T and T-T reaction rates.
A is part of something known as the astronomical S-factor, which describes all the nuclear physics of the fusion reaction. It describes the quantum physics of the nuclear binding under the strong force, and the probability that D-T or T-T nuclei fuse if they come into contact inside the Coulomb barrier.
It looks like the S-factor is something that physicists don't have a complete theory for, and hence A is measured by experiment rather than predicted using an equation in fundamental physical constants. Thus the best answer we may have is that D-T fusion is more preferable than T-T fusion just because it is.
R is the square root of something called the Gamow energy, which describes the probability of the two nuclei quantum tunnelling through the Coulomb barrier. R is similar for the D-T and T-T reactions, but the fact its in an exponential leads to a factor 1 to 5 in favour of D-T fusion reaction rate than T-T fusion.
R is related to the number of protons in each of the fusing nuclei, and the reduced mass of the nuclei. The consequence of R is the probability of tunnelling decreases rapidly as the atomic number and mass increase, and is one reason why only the lightest nuclei are real fusion candidates.
m in the equation (the reduced mass) in this case is equivalent to the mass charge ratio that you asked about. You're right, it does favour the T-T reaction over the D-T reaction, by about 10%. This is nowhere near enough to offset the differences caused by A and R however.
I didn't understand your point about why D-T fusions are 50% less probable than T-T fusions - can you elaborate?
As it happens there are some fusion scientists working down the corridor from me, so I'll ask them to explain the astronomical S factor.