Fast neutron cross scattering sections in the couple MeV range barely vary over more than the range of 1-10 barns
1-10 barns is, of course, by definition, an order of magnitude. There is a massive difference between 10 barns and 1 barn. Tenfold, to be precise.
More to the point, you can't just combine all cross sections like that. The energy imparted from an elastic collision isn't the same as from an inelastic collsiion, which isn't the same as an (n, gamma), and so forth. Elastic collisions are particularly low energy, particularly the higher Z the target. Taking them out of the equation yields much greater differences between materials in the range of a couple MeV. The upper end of the neutron energies are "somewhat" similar (up to about one order of magnitude), but down below 6 or 7 MeV or so there's quite a few orders of magnitude difference.
Likewise, total cross sections have no bearing on the accumulation of impurities in the material. The particular cross sections are relevant not only in terms of reaction rate, but also what sort of impurities you tend to accumulate and what effect they have on the properties of the material. Which of course varies greatly depending on what exactly they are.
Integration of annealing cycles into blanket design is not brought up enough in some design studies, but is a consideration to help
It's not a side issue, it's a fundamental issue to the design of a material designed for high temperature operation under a high neutron flux.
Blanket design is extremely constrained by tritium breeder ratio to ensure more tritium is produced than used, which squeezes volume allowed to be used by coolant,
... but they have much lower neutron flux to worry about. Gen 4 reactor designs are in the 500-1000 C temperature range, exceeding in some cases what is thought reasonable for fusion blanket design. ... Blanket replacement is considerably more complex than fuel replacement in a fission reactor
Perhaps they've been heading in a different direction since I was last reading on the topic, but I was under the impression that a prime blanket material under consideration was FLiBe. Which operates in a temperature range of 459-1430C, and is its own coolant. That doesn't change what the first wall has to tolerate, but as for the blanket itself, you have no "structural properties" to maintain, and cooling is only limited by the speed that you can cycle it.
The last paper I read on the subject also suggested that for breeding purposes one needs not only beryllium (they were reporting really poor results with high-Z multipliers), but the optimum ratio (to my surprise) worked out to be significantly more beryllium than lithium. So building structural elements out of beryllium serves double purpose, you don't have the excuse of "I need to use steel because it's cheaper" - you need the beryllium either way. It's strong, low density, similar melting point to steel, but retains strength better with heat, and highly thermally conductive. Beryllium swelling from helium accumulation stops at 750C+ as helium release occurs. So pairing a beryllium first wall with a FLiBe-based blanket seems like a very appropriate option.
Please don't get me wrong, I'm not at all disputing the great amount of engineering work left to do. I'm just more optimistic that appropriate solutions will be found. Perhaps I'm just naive in that regard