Let's take a quick back of the envelope look, without wishful optimism already in force.
The net energy of a U-235 fission event is approximately 235 MeV, That of a fusion event involving deuterium and tritium is approximately 18 MeV, less than one tenth of the energy of the energies involving a single pair of atoms. The slow neutron reaction used to generate tritium from lithium itself yields roughly 5 MeV, which might be possible to harvest. If we count atom by atom, rather than by mass, U-235 still yields roughly 10 times the energy of fission. But to produce enough tritium to harvest and actually fuel a fusion reactor, let's assume that we're recovering as much as 1/10 of the fission events as usable tritium fuel. That's a _very_ optimistic number, refining nuclear materials is quite dangerous and quite wasteful.
That means relative power output of the fusion plant, at the most optimistic 100% efficiency of the fusion plant itself, of 1% of the energy output of the fission based tritium source. Even a factor of 10 improvement in any step, or a few factors of 2 improvement at several stages, leaves the fusion plant far behind the energy production of the fission plants needed to fuel it.