Neural networks tend to maintain limited sustained activation to preserve functional properties, i.e. they can't convey information or compute things when completely active or inactive. This only happens within a small range of parameters. Let's then pseudoscientifically pretend that this awesome paper can provide the neurological basis for the hedonic treadmill and can be self-similary extrapolated to conscious functioning: We don't function optimally if we continually over or under react to our average stimulation, and evolution wants us to function optimally.
So to keep our functions optimized we want to be able to respond to the broadest range of inputs possible, we want to be dead center on "average stimulation" so that we can still compute and respond appropriately in events of lower or higher than usual stimulation: the hedonic set point here is your relation to your average stimulation. You might be a happy peppy person slightly overreacting to average stimulation or a lethargic person, underreacting. But it doesn't really matter, it's evolution using you to test whether that happiness point might be optimal to make you behave well and reproduce.
Let's continue assuming that this set point is achieved by a regulated process in neural networks by adapting their sensitivity to their inputs over time to stay at that optimal point.
Final assumption: if we don't optimize to our average inputs and too many of our neural networks respond too little or too much to their input for too long our brain will compute for a depressive or manic lifestyle. Both perfectly valid lifestyles, but seen by the evolutionary watch dogs as 'inferior' lifestyles. And usually reported to not be all that great. But hey, life itself is usually reported not to be all that great.
This paints the picture of the hedonic treadmill: as far as subjective well being is concerned, people tend to get used to their situation and be content, unless they become depressed or manic. That's ignoring all other things that can go wrong in the brain that aren't purely 'happiness disorders'.
This was what I believe keeps us at a point where we "behave evolutionary optimal". For evolutionary purposes however it does not matter if we consciously enjoy or detest the ride we're stuck in, as long as we see it through and procreate. That gives us some artistic freedom in the conscious experiencing of our ride. We can change our outlook on things: it is perfectly possible to be completely content with being eaten by a bear in the near future, but completely impossible to have any kind of control over your behaviour while it is happening. So when receiving negative or positive inputs we lose control over our consciousness, and evolution takes the wheel, but at our hedonic set point when things that we're used to are going on we can take control over our outlook of our existence.
So while I don't think you can, or want, to be in a permanent elevated happiness state (mania is no fun) you can permanently change how you perceive being forced by evolution to be stuck at a hedonic set point.
Intuitively you might want to resist this, or find it unfair or unwanted and it is this resistance that will create suffering while you're neurologically at a state where you could be perfectly content and at ease.
Unless we develop technology capable of changing how our neural networks function, I think nihilism (as viewed by for example Albert Camus, not online edgelords) and radical acceptance (as part of Dialectical Behavioral Therapy) might provide the most useful tools of permanently changing your outlook on your existential conditions.
To provide an answer to your question: I think in wealthy countries our best intervention to improve our subjective well-being is a continuous effort toward radical acceptance of our existential condition aided by philosophy, therapy and a healthy lifestyle.