Given your situation, as a simple approach I'd calibrate the GW1000 to a +0.3 °C offset, as the array came originally with the console and assume that the WH24/WS1002 (2) combination has the "true" temperature.
Or your trust the GW1000 (1) more.
Or your take the average of the two readings.
If you happen to have another temperature sensor (3) which you can place close to the array, you could get another outdoor temperature reading and find a different offset to apply to the console(s) based on the readings of (1), (2), avg of (1) and (2), (3), average of (1), (2), (3). Your choice.
You can also try to place the GW1000 indoor sensor close to the outdoor sensor of the array (maybe bring the outdoor array very close to the GW1000, best with no sunshine exposure, and compare its readings of outdoor temp with its readings for indoor temp. And take the average of this measurement (or add it as (4) to (1), (2) and (3) to create a new average to find an offset for each console.
On the other hand, given the accuracy of the GW1000 (Temperature range: -10°C – 60°C (14°F - 140°F); resolution: 0.1°C, or 0.1°F; accuracy: ± 1°C) and the WH24 (same accuracy), the "real" temperature, even measured with 3 to 4 different sensors and taken their average, will remain a pseudo-exact value. You are looking here for an accuracy which the sensors cannot provide: your deviation 0.3 °C, sensor accuracy 1 °C.