Its good to see there is somebody left who has not clouded their thinking with the boolean addiction and can answer the question intelligently.
My two main objection to booleans are:
Booleans often produce out of tolerance geometry or other booby traps that can trip you up later. This objection applies to most of the other solid modeling tools also.
The notion that booleans save you time is mostly an illusion (another example of cloudy thinking) When booleans fail users waste an enormous amount of time. Booleans are the biggest waste of time for those who don’t know how to model without them. Boolean users often waste time setting things up for that magical moment. If you learn to model without them you may discover that a lot of the setup operations are not needed.
If you learn to model without them you are in a position to decide for yourself if they are helping or hurting you in terms of productivity. And as an added bonus you get out from under the cloudy thinking and start to see new possibilities and avenues to getting the results you want.
It is no coincidence that new users (and some old users) are often surprised to learn that you can create more robust and accurate models without booleans. How are they to know if nobody tells them otherwise?