If you want to be a teacher or hold a research or Engineering position at a more research driven corporation, and get a confortable life and safe career, get a PhD.
If you want to work your ass off for +10 years at least, life a insecure and stressed out life, have 0,1% chance of being insanely sucessful, rich and change the world, do a Startup.
Doing a startup trying to achieve only "career recognition" or improve your resume so you can land a job at Google or Microsoft is a total waste of time. You'll likely fail, lose your money (and your investors', if you get any) and achive nothing. Being a sucesfull entrepreneur would definitely make your resume better, but that's not the end goal.
When you see startups being acquired by Google you shouldn't think the founders reached their goal. Quite the opposite. It's usually a mild failure (an "acquihire" or "softlanding"), they couldn't scale their company or raise more money and selling to Google was the best thing they could do (besides shuting down the business). Or it was a mild success, the entrepreneurs sold to Google because they would get some sizeable amount of money fast enough, rather than taking the risk to see if the company would become much larger down the road, to sell or IPO at a much higher valuation 5 or 10 years later, for instance.
In either cases the founders won't usually be happy to have become employees once more, and will leave the acquirer as soon as they can and do something else. They can't leave right after the acquisition because he will be vested or will have golden handcuffs (ie.: won't get all his money if he leaves before 2 or 3 years).
Ps.: I have a startup (for 6 years now) and before that I was studying to get a Master's Degree. However, my A plan was always to start a company even once I started the program. So once I got some money to start, I dropped out. I was already quite tired of the academic life and don't regret it. Don't regret the time I've spent doing the master's also, as it was a way to keep myself ocuppied and get some ramen money, while having a lot of free time to work on my product and think.
For me, being successful as a startup founder, is to bootstrap (I don't want to loose the freedom of making my own decisions) a business that is solving a "real" problem for enough users to have reasonable income. I am not planning to scale just for the sake of scaling or to raise fund for expanding. I am making the conscious decision to have a sustainable business with no exit strategy.
When considering a PhD, the best thing to do is look at the advisor(s) you'd be working for and their repertoire of projects. Looking at the "discipline" isn't precise enough; there's a wide range of stuff happening within fields like information systems or computer science, everything from very theoretical to very practical.
Thank you for the advise, I agree that the discipline is not enough to make the decision.