As Nunoxic noted, doing research as a part of your job is possible. Also, doing research at a graduate program that isn't at a top school is possible. But doing research on your own, and hoping to bring innovations to a field... is highly unlikely to work, in my opinion.
My opinion is based on experience with a bad schooling system (undergrad and Masters in Serbia) where publication is mostly irrelevant to survival of the teaching staff. Access to research journals is also severely limited, and, as a consequence, access to international conferences, international collaboration and all that. Essentially, supervisors would offer minimal guidance, and students are always free to choose their research topic. This is still more guidance than you would have on your own.
Then I got myself into good schooling system (a second Masters - to catch up - and now my PhD studies), in The Netherlands. Neither of the two Dutch universities I've been studying at are top schools in the Shanghai sense: their rankings fall between 100 and 150. Still, in comparison with Serbia, the difference in what I've learned and achieved since being at these universities is staggering.
- I don't spend months of my time sifting through articles in order to slowly begin comprehending which papers/names are the most relevant ones in the field.
- I have someone to immediately tell me if a research question is worth pursuing.
- I have someone to immediately point out similar research.
- I have someone to immediately correct small mistakes that would prove important at the end of data collection
- I meet important people in the field, and get feedback from them, with no cost to myself
- I work with technology that costs a great deal of money, with no cost to myself
The essence boils down to having access to people who know what my work is about. This speeds up my understanding of the subject matter by an order of years. Science is so fast-paced nowadays that it's very difficult to make breakthroughs without collaboration.
In sum, my advice for someone in your position would be to give priority to non-top graduate schools, if it doesn't work put with top ones, instead of trying to be a solo researcher.