Why 1 Watt R's for PI Bias and Tail ?

[PCollen]

Looking through the Express BOM, I see 1W R's called out for the PI bias and tail . Do voltage drop or current draw measurements/calculations bear this out as necessary, or is this just for historical purposes. To dissipate a half-watt, the 10K tail would have to drop 70 volts and the 470 about 15 v. Given that the quiescent current draw through these components is not even close to that necessary to give these voltages, and the NET AC current draw through these components is theoretically zero, I can't figure out why 1 W would be necessary and that 1/2 W would be more than sufficient.

[FYL]

KF used large CF resistors there because he had them handy and they looked nice.

Ditto for the skyscraper-mounted 2x 100K/3W or the 2x 9K1/2W between filtering nodes.

BTW, using suitably derated components is a Good Thing™ when it comes to noise and reliability, so, yes, 1W components are justified even when they dissipate only 1/8W or so.