You are absolutely correct. Moreover, reifications can be useful or harmful - purely based on how they are being used. That's why maps are actually useful, except for several individuals that died in Australia and other places by trusting their navigators more than their own eyes and actual surroundings. I've been in situations where GPS malfunctioned and when I quickly realized it I understood that I should not follow the map.

Well, not any more than chairs :P

I'd say happiness exists exactly as much as chairs, in both cases we are classifying what is, in reality, just a collection of atoms, either as a chair or as a happy person, based on some external measurments

No, it's not my logic and your conclusion is incorrect. Measurement units don't actually exist (meters, seconds, inches, pounds, liters) - yet that's how we measure things and very successfully.

Abstractions are critical to human thinking in science. But not only reification is very different, you can't measure happiness even if you want to. You'll always have to go 'happiness as defined ...' or resort to people self reporting their feelings. Which is fine - I always ask my kids how they feel. I just understand that not only kids and adults may change how they feel next hour but that those feelings are unreliable. If you have kids of your own, you know how many 'tragedies' they lived through by the age of 6.

Maybe, maybe not - you'd have to define which complex set of chemical combination is happiness and which isn't. But why? We know a lot about neurochemistry. We know a lot about receptors, neuromodulators and their interactions (serotonin, addiction). As a biologist I don't understand - why would you want even study something that doesn't exist (happiness), when you can study specific parts of neurochemistry that are more practical and may have actual benefits (like reduction of addiction or help with schizophrenia)