170V, 180V, 100K, 82K, 20%, 5%......
Bah. This isn't a moon rocket, where 10km means you miss the moon completely. It will play well at 150V or 200V. It will play well with 100K+100K in PI plates, though theory suggests a slight unbalance roughly correctable if the near-side load is ~~20% low.
Resistors used to be made like bricks: a pot of clay with a handful of sand for bricks or coal-dust for resistors. One brick may be 20% softer than another, do you really care as long as the house stays up? Old radios used the cheapest 20% tolerances, most guitar amps are from days when 10% was most available, today 5% are cheap and 2% is cheep-enuff.
But 2% is not "better". The designer will assume some tube values (which really vary +/-20% or more) then do his math. If he had that specific assumed tube, the perfect resistor might be 113,456 ohms. However anything from 75k to 150K works too, and so close you may be unable to tell a difference.
Where Sluckey had a missing band, he meant "use your your cheapest or favorite color". When Silver (10%) was cheapest, that's what we had. Gold is prettier and today you can't find Silver as a commodity resistor, so use Gold. Or if 2% are being dumped cheap, use 'em; but don't think that 98K-102K is "better" when the amp will work a treat with 95K-105K, 90K-110K, even 75K-150K.
And a different designer might have favored 200K. 270K and 470K were favored in radio audio amps: you lose a hair of treble but get a bit more gain below 4KHz and use less power (cheaper filtering). Fender did that, then moved to 100K. He may be right; this is the most-copied value on the planet. But Gibson and Ampeg sold a bunch of amps with other values, and using other values is a popular tweek.