There is lots of good input here, but here's another take on the subject. The Navy uses epoxies all the time for bonding aircraft together, for coating flight decks, lining fuel tanks, coating sensors, bonding sonar arrays..etc., so theres nothing wrong in concept with the material. Epoxies are typically not bothered by salt spray and they can be very resistant to solvents like jet fuel. They are not going to conduct heat as well as metal regardless of filler. However, this might not have any significant impact on the heat sink performance. If the bond is thin and the area large the difference between brazing and thermal epoxy bonding could be so small that it's difficult to measure. You can easily calculate this using the resistances refered to elsewhere in this forum and add in the thermal resistance of the fin/air interface. There are two caveats however: 1)in very high rates of forced convection the fin to base resistance will increase as a percentage of the overall heat sink thermal resistance and 2.)Don't use the glue above the glass transition temperature of the adhesive since all the good properties like chemical resistance and mechanical strength change significantly.
In a heat sink