Source?

It's that so they can fiddle their taxes?

Why [else] would you not want to get the same money, on average, as a wage instead of having a high per diem variance?

Because, for one thing, absent tipping, pay would probably tend to be more egalitarian across front of house and back of house. And some servers (especially young females working busy shifts) would definitely lose out with a switch to a straight salary. This has been studied. Might check out the Freakonomics episodes on the topic. But restaurants that have tried to make the switch have largely gone back.

The number of attractive waitresses as proportion of all restaurant workers has to be a minority, surely? I can understand that a few will miss out on the possibility of flirting to get a higher wage, but that wasn't the OC's assertion.

Also, aren't most young women feminists, they should be against such things ...

The vast majority of "no tipping" policy restaurants I've read about fail in large part at the protestations of the staff. Maybe there's a lot I haven't read about, but I encourage you to look at the case histories. This does not mean however that it could not be accomplished by fiat as state policy, albeit with significant grumbling I'd assume.

In practice, if you literally replace a 20% tip with a 20% service charge, gross wages for staff decline significantly -- this is a consistent pattern. This strongly suggests that they are not equivalent from the perspective of customer psychology in some important ways, changing customer behavior. Many restaurants in Seattle have tried this and it plays out the same way every time.

The consensus theory among my industry friends is that Americans are more generous with their spending when there is a perception that the money is going to a real person -- tipping -- whereas a service charge goes to the restaurant company. Customers attribute the cost of the meal differently in these two scenarios and they are not as willing to spend more money if it goes to some faceless company instead of a person they interact with.

Could you cite a study or other statistical evidence? I'm surprised to find that the median server gets tips >20% averaged over all customers as a voluntary contribution. It would be interesting to see a break down of the range of tips send which restaurants were in the study, etc..

Presumably other areas of devices, like clothing or general retail are getting demands from unions to move over to voluntary tipping to tap that generosity?

In UK it's relatively common to tip some hotel workers and taxi drivers -- do USA hotel room cleaners also benefit from this effect of tips>>charges?

First reason is that everybody likes to believe they are exceptional and will earn above average tips. Second reason is they want to evade taxes.