Are "independent" scientists vetted out of the process if they're shown to have a vested interest in specific topics...such as their university receiving grants on studies of such issues that they have regulatory powers over? A conflict of interest can arise in many different ways...not just through private corporations.
The SAB does not have regulatory power. The SAB basically weighs scientific merit of different memos or provides targeted feedback, review, and research as requested. It is a panel of experts, people that serve in a variety of fields, and is made up of a bunch of sub-committees that tackle more targeted issues.
There is no conflict of interest from a scientist just because someone at their university has a grant for something. That's not how grants work, really. For instance, another lab in this building has a grant from USDA. To accuse me of bias in regards to, say, crop science research because some people in this building receive funding from the USDA is really silly. Plus, one big thing in science to keep in mind is the process, while not perfect, actively awards people that buck the status quo or find alternative hypotheses to traditional ideas.
Now, if I was receiving a grant to study, say, harmful effects of high fructose corn syurp, if I was to research that subject, publish on it, and gain respect in the field, I would be considered an expert on it. Now, you might not wanting me running the EPA, but certainly, wouldn't you want someone that actually knows the research in the field to comment if the EPA prepares a draft on how to handle, say, something related to high fructose corn syrup? You don't want me handling it alone, no - but if you were to gather five or ten of my colleagues that are working in the field that have made significant impact (easy to measure with publication scores), then you might consider yourself to have an expert body to review scientific evidence.
Maybe some of those people have a connection to industry(and they would, in this example). Maybe they don't. ALL of them would have had some connection to the field and received funding in it from one source or another at one point or another. But to declare them all unable to weigh the potential merits of something because they received funding to study the topic at hand would suggest that the entire scientific process is untenable, unless you want to just give every single scientist a flat sum of money and completely remove the aspect of sciene that relates to how performance influences funding.
Note, I am not saying individual scientists are not without bias. This is why peer review is so important... at every level. Which is a big part of the SAB's functionality.