Embedded in that is the premise that increasing the number of working hours is a good thing. If the law insists upon a maximum of 30 hours per week, then the number of hours that one is willing to work is no longer a way to distinguish between employees.
This thread started with the observation, quoting bato, "as work gets "easier" the way forward would be to reduce hours worked, not increase them".
If you build into your analysis the assumption that working more is better, then of course you'll end up with that conclusion that working more is better.
You believe there is a negotiation lever by not having a cap in the law. To start with, I would rather have strong unions able to negotiate the cap as appropriate for the given trade. Failing that though, the US has laws limiting the number of hours to work, such as the Libby Zion Law in New York, which limits the amount of resident physicians' work, and hours-of-service rules for truck drivers.
Regarding the maid example, I believe you are suggesting that it's rare enough that it can be ignored for purposes of economic analysis. My suggestion is quite different - your justification says that we should not do things that reduce the GDP. I question the primacy of that argument. We make policy decisions to have our country more in the way we want it to be. GDP is easy to measure. That doesn't mean it's the right metric, or even a good rough gauge.
If we wanted a country where people had more time for personal enrichment, then we would have a lower GDP. So what? Studying French poetry of the 1800s or building sand castles on the beach or watching football games are cheap.
And if you're worried about the poor needing to catch up to the rich, then increase taxes.