About 5 years ago I developed my own version of the algorithm, which produces largely the same results as the official poll (I *think* the small number of discrepancies are due to a couple of typos in the raw ballots displayed on the site, though I'm not 100% sure). I have used my implementation to run an unofficial poll on Alton Towers fansites (links to my own 2011-13 polls), so I have some experience of running the poll, albeit on a much smaller scale. Armed with this I was tempted to try to "go it alone" last year, but due to my lack of web knowledge I didn't attempt it.
After the steel poll skipped a year in 2011 I updated the ballots for 2012 and 13 (sorry for any errors & omissions ), and it was a major undertaking, one which I don't really have time to take on. To see the scale of the task, rcdb now lists over 3500 steel rides, compared to just 177 woodies worldwide. If the poll is to return, it may be necessary to seriously rethink the eligibility criteria for the steel ballot to ensure it is manageable yet representative. As a rough guide, I used "more substantial than a Vekoma Junior" as my main requirement for inclusion in 2013, which resulted in a 450 line ballot.
To anyone brave enough to take on the task, good luck creating the ballot! I'm sure there are loads of people out there hoping it'll come back in some form or another, as the methodology makes it arguably the most accurate coaster poll out there.