Well, sure, but why should any of us pressure google into helping people in general to keep this stuff secret from insurers, then? It's not like we'll get them to only keep our own information private; this going to apply to all of the insurance company's customers, and we (like all their other customers) will be hit by the globally higher premiums.
I can see why the subset of people who are currently exploiting this information asymmetry for profit (i.e. people with above-average risks who can keep that fact secret) would like to pressure google for more confidentiality about this, but everyone else shouldn't.

It's kinda like saying "I like to sell people lemons, so google should really not return any results related to the state of specific used cars, because that hurts my special interest." Lemon seller would like that, pretty much everyone else wouldn't, so why would people in general possibly support that kind of thing?

And just as a reminder, none of this touches concerns about keeping data from the government or employers, it's just about the "keep insurers in the dark" part. The other concerns would have to be addressed seperately, and I happen to think that keeping personal data out of the government's hands is in fact, a good idea.

(2) Things we wouldn't want our employers, potential insurance companies, or dictatorial governments to know about.

It sounds like the Google guy is throwing out the baby with the bathwater. He might have a point about (1), but his comment also seems to dismiss (2), and that's a real problem.

I'm with you on the governments and mostly on the employers (though one could argue that the latter are less important, since the market would tend to weed out inefficient discrimination).

But ... why insurance companies? The idea of risk insurance is that you pay a large entity some money so as to put an upper bound on the worst things that might happen to you - i.e. there's some low-probability high-damage risks in your life (or venture), and you'd like to reduce the variance from those events. If you go into an insurance contract knowing something about the probability of the risk that the insurer doesn't, you're just trying to rip them off. And the insurance company will know about that possibility, adjust for it, and as a result demand higher premiums from all their customers, because their ability to tell the ones that are hiding information from the ones that aren't is imperfect. That ends up with people with a normal risk that want to buy a policy subsidizing those people that know privately that they have an elevated risk level. And of course then that leads to second order effects, with that kind of system attracting people with elevated risks, leading to higher premiums for everyone, etc.

Insurance works better if the insurance company has a good understanding of the risks they're actually insuring people against. Unilaterally withholding information of them can make individual sense, but is ultimately a negative-sum game.

hidden cameras in the voting booth; including those built into cellphones held by the voter [because the voter could be trying to prove what (s)he voted [in order to sell her/his vote or avoid retaliation if s/he did not vote as instructed].

Trivial to fix: Hand out as many ballots to each voter as they ask for, but only allow them to drop one envelope into the urn. They can make as many fake votes as they want, and photograph them however they like, but there's no way for them to prove they really dropped whatever they photographed into the urn, as opposed to throwing it away and filling out another ballot with their real vote.

My suspicion is that with common use of these exosceletons the percentages of groups that 'need' them will increase. The trend, of course, could be reversed if people took more care of the standard implementation of body functions. However, that would not create another growing revenue stream in the health care sector.

<sarcasm>Yes, and by the same token we should disallow the use of airplanes cars and bicycles for people who are physically capable of travelling the distance under their own power, no matter the cost or effort.</sarcasm>

What would be so bad about the scenario you describe?
Taking good care of human muscles is a time-consuming, high-effort activity that many people don't find especially pleasurable, and that in fact isn't worth the results according to their own judgement. Historically people have increasingly offloaded activities previously done by human muscles to machines, freeing up time and energy to do other, more productive or pleasurable things.

I'd want one of those suits. Well, not right now, because right now they're still very expensive and likely error prone, but I'd definitely want one for daily use once they're in mass production and the major problems have been worked out. And not just for the sheer nerdiness.
Who are you to decide for people what they should and shouldn't offload to machines? If these things get good and cheap enough, they will be used by a large part of the population, who will be happier and more productive as a result. The people who use them will happily pay the price to make and maintain the machines, just as we currently pay the price to do the same with cars. And we'll all be better off as a result.

It is, for instance, perfectly possible to replace all != with is not.

That's completely wrong. is not tests for identity, != tests for equality. For instance:
>>> True != 1
False
>>> True is not 1
True

Personally, I also think 'is not' is a horrible name for a python operator. Without this operator, the expression 'X is not Y' would still be valid Python, with a meaning equivalent to 'X is (not Y)' - which has a different meaning from 'X is not Y' in Python. Therefore, 'is not' could well be confusing to people who don't know it, but do know the other python operators. IMO, 'isnot' would have been a much better choice.

I imagine that the DECT link would just be carrying a nice SSL session, and it wouldn't much matter.

<sarcasm>Umm, yeah, that would work great.</sarcasm> Less facetiously, it would indeed have been possible to make things secure using SSL; but just using it isn't enough, at least as long as you use standard certificate authorities for auth. If you make your own, this particular crack probably wouldn't hit you.

How much is it going to cost per month or year to have a public ipv6 address. You can't say nothing because they will have to be allocated by someone in some way that not only tells routers where to direct stuff but to ensure that your traffic in LA isn't being routed to the same IP in Bermuda or Russia.

Well, right at this moment it costs ... let me check now ... nothing. If you don't believe me, check offers from existing ipv6 tunnel endpoint providers. sixxs, for instance, will offer you a (static) /48 network (that's 2**80 addresses, or more than the square of the entire ipv4 address space) for free. I can attest from personal experience that this works and is totally free as in beer, since I've been using their service for over two years without ever paying for it.

This is just an ipv6-through-ipv4 tunnel offer, so you do need working ipv4 connectivity to use it. Actual native ipv6 connectivity usually isn't gratis - but then the same applies for ipv4 uplinks.
Global ipv4 addresses may be expensive, but ipv6 ones really aren't, and unless ISPs start introducing completely artificial scarcity on that front (which they might, being greedy businesses and all) they're not going to become scarce in the forseeable future.