Hi Steve,
Two notes on the W3(OH) observation presented above: because W3(OH) is at a relatively high declination (~62 degrees), the transit time was a bit longer. Total integration time was about 15 minutes. Furthermore, the system temperature was ~100 K, as estimated from hot- cold measurement. Finally, the W3(OH) mase feature was quite broad, so I could get away with using 35 KHz bandwidth. If you plug in these numbers in the radiometer equation, you get a minimum detectable flux of 626 Jy (5 sigma).
By the way, it should be possible to increase the integration time simply by averaging transits from multiple days. Most masers vary on timescales of a few weeks to a few years, so a time resolution of ~1 week is perhaps good enough.
If we assume a 1.2 metre dish, Tsys of 150 K (maybe a bit pessimistic), 8.2 KHz bandwidth, and 7 days of transits averaged (49 minutes integration time), the detection sensitivity is already brought down from 1970 Jy to 746 Jy... So the strongest methanol masers could be within reach of a 1.2 metre dish in transit mode if you average multiple days of transits.