None of those schemes for neutron shielding substantially affect the mass budget of a p-¹¹B reactor. Neutron shielding is not the issue. Any shield that can knock out the bremsstrahlung, never mind the gammas, will likely be most of the shield mass regardless of how neutrons are handled. Depleted uranium is probably better than lead for the bremsstrahlung, but it doesn't seem to offer much advantage, if any, for gammas.
Assuming the gammas are actually a problem. I'm still hoping...
The x-ray deflection stuff
is interesting, as I commented three posts down on your second link.
Mostly, though, MSimon is right. A foot of lead on a sphere with a 5 m radius is 1155 tonnes. Reduce the radius to 3 m and you get 433 tonnes. Two metres? 202 tonnes. Shadow shielding gets similar bonuses, assuming the wiffleball and reactor scale at the same factor (which is probably not the case; the wiffleball probably scales faster because the conversion system has design parameters that don't scale).
Also, if you can get 900 T out of this stuff, that gives you (using MSimon's
estimation technique) 90 T at a 1 m bore. Assuming the 100 MW reactor was supposed to be 10 T at a 2 m bore, that's 82 GW. Overkill, and probably impossible to design compact collectors for, but at least we have some headroom... even at 500 T you get almost 8 GW with a 1 m bore...
...now, what's the critical current density in one of these?