It's a good question. This past week, I was re-reading some of Jay Jaffe's work on the ballot over at FanGraphs. And on his ballot he did vote for Sosa (and Abreu, since we're talking RF), and he explained why, not only pointing out the enormous home run total but also the comparisons against other right fielders and hitters linked to PEDs. Like Bonds and Clemens, he was tied to PEDs before there was testing in place in the majors and before there was any punishment, unlike when Ortiz, Rodriguez, and Ramirez played their careers. There are a few things that to me seem to be a drag on Sosa:
-- Voters who aren't going to vote on that era are going to leave him off.
-- The counting stats of hits, wins, homers, etc., don't carry the same guarantee in this era of voting because there are more homers, more members of the 500+ club, and also the advanced metrics are giving new looks, new rates, new ways to view careers. It's not quite wins for starters, but it puts single-dimension candidates in a larger context.
-- His vote total seems stagnant, so with a 10-person limit there's a choice to make about supporting someone else who might see growth in support or need precious votes to remain in the conversation.
-- He just doesn't rise to the level of the other candidates for some. I mentioned oWAR earlier and listened some of the candidates. Might surprise you to know that he has a oWAR of 50.4. Again, I'm not saying that's the be-all, end-all stat, but it's an informative stat when it comes to how much a player contributed offensively. Sosa played 2,354 games and had an oWAR less than Rolen's 52.8 in 2,038. And earlier in the chat you can see how that compares to others on the same ballot.