I know nothing about this specific system, but based on general priciples you could consider an increased contribution of solid solution strengthening of Ti in W. If you have excess Ti, or you have somehow not reached phase equilibrium, there could be a trasformation where bcc W sucks up the available Ti with increasing temperature until there is nothing left. Past this point, at higher temperatures, normal temperature-dependent softening is observed. Since you observe pools of Ti in your microstructure this mechanizm could be considered.

Thank you Pål Baggethun

Tungsten alloys are very brittle at room temperature. When the temperature rises, the brittleness slowly decreases while the strength remains (only above 900-1000°C does the strength of the tungsten begin to decrease). For this reason, the strength measurement at temperatures between 700 and 900°C brings higher values than the room temperature measurement. By the way, the same thing happens with other brittle high-temperature alloys, for example with chromium (measurement at 400-500°C gives higher strength values for chromium than measurement at room temperature, because chromium no longer becomes completely brittle from approx. 300°C).