"Least signaling" is a bit too vague to work with. Under your definition, the only criteria seems to be the degree to which we can talk about it. Anything at all that is socially acceptable to talk about can be construed as a signaling activity, so this definition is not very useful compared to say, basing it off of how many people would do it being observed vs. not observed, or which country they would choose to live in in the data you cited.

If it's pleasant, I can brag about it to show how much leisure time I have / how my life is better than yours. If it's unpleasant but useful, I can use it to show how good of a work ethic I have that I got it done. If it's expensive, it shows I'm richer than you. If it's cheap, it shows I'm thriftier than you. If it's beautiful, it shows I have better taste than you. If it's ugly, it shows that I'm enlightened enough to know that it's not really ugly, or I'm thrifty, or I'm above aesthetics. If it's a complete waste of time, well, it lends itself to a good story and lets me showcase my sense of humor. If it's really, really unpleasant and pointless (like a serious disease/injury), I can gain your sympathy, improve my image (who speaks ill of the dying?) and make people remember me better than they otherwise would. If it's taboo, I probably don't talk about it, but if I could, oh, you bet it'd be a signal of some kind.

This seems to be right on the brink of the Perfectly General Explanation, if our only criteria is "Can we find a way to signal status/fitness with it?." The fact that damn near everything we do can be used to signal does not mean that it is actually done for the purpose of signaling, or indeed that signaling affects it in any way.

Moreover, 11% of people are willing to have their car get broken into as long as the people around them get their cars broken into more. Same thing with being sick. Thus, the argument that someone, somewhere, could construe getting AIDS as being mostly positional (e.g. he gets AIDS for signaling purposes) does not mean that getting AIDS can be seen as being even partly positional in the vast majority of cases.

You need some kind of definition for "signaling activity" such that it doesn't contain every direction and magnitude of pretty much everything we can possibly do. Using the percentages in the data you posted might work. Examining how much behaviour changes in the absence/presence of observation would also work, though it's harder. But the current discussion does not seem productive; all it does is show that under certain circumstances, if we can't talk about certain things, we won't signal with them.