These polls are flawed statistical data due to nonresponse bias. Considering from what you say that these polls ended up 56/44 for and against RDF I’d say accounting for nonresponse bias in those polls the true preference of the majority of players doesn’t look good for the for RDF side.
Getting a group together might seem daunting for retail heads, but it’s really not this boogeyman that the for RDF side is trying to make out it is.
If they do choose to implement a form of RDF they could do so while keeping a likely large portion of the against RDF group somewhat happy by doing something like this compromise.