I agree, better cooling is not going to be revolutionary, would you go for an incremental improvement?
The alumina is in a binder matrix, should not damage the wire.

Mixing compounds
The results form working on conductive epoxies and other thermal compounds for IC packages their effects are compared by their mass ratio.
Lets use alumina at 40 and 90% of the mass.
Epoxy at 0.35 at 10% mass.
Expect to fill mold with alumina and epoxy is only a binder. (epoxy fills the voids only)
I believe the mix is calculated like parallel resistors.
Result will be 3.2, almost 10x improvement over just epoxy.
And is twice that of Mortar:-)
The mix should be stronger than epoxy, allowing a thinner structure or Ribs.
Thermo conductivity of 3 is not great, 10x better than 0.3!

Since a filler saves $ selecting the right one can offer other benefits.
Most of the other fillers discussed offer none that I can see.

ATH may offer benefits, after searching for two hours still do not know its thermo conductivity. Considering its used as a fire retardant I would be surprised its near alumina's.
So went with something I could! (and was available and cheap)

Designed many power electronics circuits, during a fault the power device would fry.
By thickening coper and increasing area allowed it to survive, never was a good heat sink, just good enough!