wvmarle:
No. Normally in case of no communication, a line is left floating - and the pull-up resistor will pull it high. Then to communicate, the line is pulled low by the sensor as needed. Examples of this include OneWire, I2C (aka TwoWire), and the DHT11/DHT22 communication. Other protocols are driven high and low, such as SPI, and don't need a pull-up or pull-down resistor.Now why don't you start to provide some real information on this anemometer, such as a link to its data sheet (a quick Google search on the type number you provided didn't give useful results), the circuit diagram of your setup, and your code. The first part is actually the most important. Without data sheet it's basically guess work. Who says that the blog posting you refer to is even correct?
I think there is no circuit diagram available for TX20, and that's why I am guessing, why pull up resistors work. La Crosse WS2300 weather stations use TX20 and all its sensors are "closed" source, as almost all of commercial weather stations.
But could my theory explain the behavior? This sensor doens't use I2C or SPI, and mayby its designers just decide to use weird line logic?