Asklemmy

You have asked the most important question in this topic. Privacy and security only have meaning when you develop a threat model or encounter a threat. With digital security it is usually pretty straightforward in that you don't want anyone else controlling your computer or phone and using it for their own ends. And a lapse in digital security can ruin attempts to secure privacy.

Privacy is where threat models should be developed so that you (1) don't waste time worrying about and working around nonexistent threats and (2) can think holistically about a given threat and not believe in a false means of privacy.

For example, if you are of a marginalized community, closeted, and in a very unsafe living situation, your main threat model might be getting doxxed and outed. To prevent this you should ensure that there is zero to no information that would link your real identity to an online identity and you should roll accounts to ensure small slipups can't be correlated. VPNs probably don't help in this threat model but they don't hurt either. A private browser does nothing in this situation. Securing your phone and not leaving it unlocked anywhere is good for this situation (sometimes privacy isn't really about tech but behavior). Using strong passwords that can't be guessed helps with this situation. Making a plan to move to a safe living situation so you can be out will resolve the threat entirely, though it may mean needing to think about new ones.

Notice that the government was not in this threat model and that it was more about violence towards the marginalized. Cis white guy techbros generally have nothing to worry about re: infosec and are just being enthusiasts or LARPers. Nobody is showing up at their house with a gun and the feds are not going to arrest you for having the most "centrist" political takes and actions available. The people that need to project themselves are those facing overt targeted marginslization or who take political action that the government wants to, or would eventually want to, suppress. For example, the US government labelled anti-apartheid groups as terrorist organizations and intimidated or jailed those they could identify. It has a habit of doing this to any advocacy groups that gain steam and actually pose a political threat to their opponents.

Even if you don't have a threat model, though, having good digital hygiene is useful in case one develops in the future. You may currently do political work that seems safe, and it is because it is not perceived as a threat. Let's say you help organize unions. But there have been times where organizing unions would mean you're targeted by the government and hired thugs and those times can easily return. If they have compiled a database of likely union sympathizers, will your name be in there? Maybe that's a risk that you just take. But maybe you should use good privacy practices so that you can go underground when needed.

The latter applies to the threatless cis white techbro "centrists". Such an individually may someday change politically or in their gender identity and having good practices would then pay off.