Unfortunately, I think we're passed that point. We need all of it. We need to change our society in a much more fundamental way than most are willing to accept. It's not "just" getting rid of aviation and meat, it's much more. We need to rethink society, and we need to do it now.
(I say this as a vegetarian with no kids, who doesn't know how to drive and has spent the last few years working in the public transport sector; knowing none of these small actions will offset even a tiny fraction of the destruction being inflicted by oil companies, their paid-for politicians, etc.)
Which means there is one solution: we need to reduce drastically our dependence on energy. Everywhere. It's not a simple optimization problem where you can "profile" and solve the biggest problems in order. The only way this works is if we start fundamentally changing society.
It's probably easier to go the other way: identify the largest problems, and try to tackle those. No point trying to fight hundreds of entrenched behaviours and industries to prevent only a fraction of a percent of emissions. Better to make a dent in the big boys (electricity generation, transportation, land use, etc.)
> Starlink, very cool tech, is it really required? I'd say no.
It's certainly debatable. A handful of geostationary satellites can provide basic connectivity; starlink's only purpose is speed.
> Another one is driving the kids to soccer practice in an ICE powered SUV, absolute lunacy given where we are with the climate crisis and a great way to contribute to a shit future for your children.
Right: this is higher-up the list of things that can make a difference. It's only lunacy from a distance: it's often a local optimum, due to problems like car-dependency, suburban sprawl, low fossil fuel prices, gameable emission standards, oil subsidies, under-funded mass transit, vehicle-size arms race, etc. Improving those will change the cost landscape, and give local optima that are less pathological (e.g. kids transporting themselves to soccer practice).