Let's assume he really holds 500 patents, all of which have a contribution from Roohparvar. In academia, there are plenty of professors who will publish articles upon articles just based on one project. They just take bits and pieces and publish them as opposed to other professors who publish an entire project at once. What's stopping Roohparvar from doing the same with patents, i.e., patenting bits and pieces of a larger system?
Nothing except money. Since it costs quite a lot of money to push a patent through the system, generally you will want to do just the opposite: made your patent as BROAD as possible, and therefore more valuable. There is really no prize for largest number of patents (despite the fact that the company I used to work for, IBM, repeatedly touted their leadership in annual number of patents issued. But at IBM, I can tell you that the IP lawyers were not trying to intentionally break up inventions into dozens of different patents. On the contrary, the lawyers would get you to expand your claims as much as possible to cover every foreseeable application space the single patent would cover.
The other issue with Frankie's patents, as others have theorized, is that he simply attached his name to patents where the inventions were actually his underlings. While it's certainly possible he did this, it is also very illegal. Every inventor listed on a patent must have contributed some novel intellectual property to the invention. IBM was/is very serious about this as well. During the review process, each inventor is required to describe their specific contribution.
I am SURE that not all companies operate like this. And it would be extremely difficult to prove after the fact, but I believe finding inventors listed on patents that did not actually contribute to the invention would be a valid reason to nullify the patent. So it's a very risky way to do business. CEO's that add their name to the patent without actually contributing to it are asking for trouble.
Now I do have several friends that do indeed have 100's of patents legitimately amassed over the years, with the strict requirements I listed above. They do have a process they use. In the most successful case, the guy held a weekly breakfast meeting where we all sat around and brainstormed ideas. Usually there were 10-15 of us. Sometimes we would hit upon an idea that seemed patentable and 2-3 guys would develop the idea and meet separately. The leader would make sure he attended all these separate meetings and he would generally contribute one small nugget towards the invention so he could get his name on it. In this way he kept his patent pipeline full with 2-3 patents applications a month (and after so many patents, he was an expert in determining whether something was patentable or not and how to write it up). He did have to work hard...he did most of the writeups on his own time in the evenings and he spends a fair amount of time developing these ideas. Because IBM valued their number of patents issued, they do allow (even expect) inventors such as him to spend 10-25% of their time developing patentable ideas.