PhDs in IT do not pay off, at least salary-wise. The time is takes to complete them is much better invested in a regular job, gaining practical experience.
I like both. Industry is more pleasing because you work on products that are directly used and you are judged (at least in a small company) on tangible results. The downside is that work is mostly driven by the most urgent needs or cash flow, so there is far less time to explore certain things deeper.
In academia on the other times, you are judged on getting grants (which is definitely process that is to a large extend out of your hands, and tends to favor people who were already successful in getting grants). On the other hand, you have far more time to explore, try, fail, and try again (as long as you are not a professor).
One should also realize that graduate studies highly differ per country. E.g. I did my PhD in The Netherlands, and there are no tests, or anything alike. I basically got funding for four years to do research, a small amount of teaching and to write my thesis. There was only one serious sanity check after the first year to filter out PhD candidates who were very unlikely to complete a thesis in four years.
I think this is a point that deserves more emphasis. It's easy to forget the degree to which US education is uncommonly expensive, among other things that skew its ROI.
Makes me wonder if I could afford a CS education abroad (being self-taught), without taking loans.
So, income-wise it also gives you an advantage of four years.