Earlier quoted context omitted.
> If given no alternative choice that is evil-free, I will chose the least evil option as my vote. I was on Android until the start of 2019 when my wife bought a new iPhone and I took her 6S. I'd had enough of Android spying on me. My argument was the same as yours: Apple don't make (that much) money off my data so they MUST be better than Google. Right? When I think about it, I honestly cannot say that they have bet…
I don't think services have to be open source for us to trust their privacy. In the context of Apple there has been enough conflict in the press with the FBI and other authorities that their actions speak enough I can trust them in the absence of source code. And I think there is enough financial strategy for other companies - such as Samsung, who for example sell smart TVs at a loss to recoup money from Ad/tracking…
I agree that we don't NEED open source but it would definitely help.
In all honesty I'm not sure any corporation could fully reassure everyone that they are trustworthy but being open source would at least show they are serious about it: they may have a back door but there's a risk it will be spotted!
That being said (and is pointed out by another commenter here) Microsoft are showing that on one hand they can giveth (open sourcing .NET) and on the other they taketh away (the incessant forced telemetry).
Now that I write this I believe the answer is something like the Librem 5 with self-hosted backend infrastructure (I'm not a employee but I will be buying one at the start of 2020 when my current contract expires - I wish I had backed them at the start :()... it's not ideal but I think it's the only solution going forward.
On a slight tangent, there was an experiment done with bookies and laymen to predict winners based on 5 points of data then scaled up to around 80 points of data and it showed that the bookies were no more accurate than the laymen regardless of how much data they had at their disposal (my Google-fu is not working today so I can't find the experiment and my figures may be a bit off).
I've asked on here before for a link to research that shows more data moves the needle enough to justify its collection... At the time of writing this, I do not believe it does.