Here are some current favorable-unfavorable poll numbers (CBS News/New York Times Poll, data collected May 18-23, 2007):
Hillary's favorable/unfavorable/undecided/unknown numbers are 38% / 42% / 19% / 2% (
http://www.pollingreport.com/C.htm)
Obama's favorable/unfavorable/undecided/unknown numbers are 34% / 21% / 25% / 20% (
http://www.pollingreport.com/l-o.htm)
Edwards's favorable/unfavorable/undecided/unknown numbers are 30% / 28% / 27% / 14% (
http://www.pollingreport.com/E-F.htm)
It is true (as you say endlessly) that Hillary's unfavorable numbers are boosted by her high name identification numbers. Only 2% of the population sees Hillary as an unknown quantity (not to be confused with the 19% who haven't made up their mind about her). In contrast 20% see Obama as an unknown quantity.
As compared to Hillary, an additional 18% of the population don't know Obama. If the "extra 18%" who don't know enough about Obama to rate him got to know him better, and if they are typical, we could expect that about 21% of that 18% (or not quite 4%) would have an unfavorable rating of Obama.
It is true that Hillary's unfavorable numbers are "boosted" by almost 4% as compared to Obama's because more people know her.
However, the MyDD analysis is facile (and inaccurate) because if the "extra 18%" who don't know enough about Obama to rate him got to know him better, and if they are typical, we could expect that about 35% of that 18% (or just over 6%) would have a favorable rating of Obama.
Just as Hillary's "negative" poll numbers are boosted less than 4% (relative to Obama's) by her high name identification, her favorable ratings are also boosted relative to Obama's by more than 6%.
The main problem isn't Hillary's raw unfavorability numbers (42% in the CBS News/New York Times Poll). Her main problem is the ratio of favorable (38% favorable to 42% unfavorable).
The raw number which is troubling is the low unknown factor (just 2%) and the low undecided number (just 19%). This means that Hillary does not have a huge population of potential new supporters who have not tentatively made their minds up about her. In contrast, both Obama and Edwards have a population MORE THAN TWICE AS BIG of potential new supporters who have not tentatively made their minds up about them.
I AM NOT ONE WHO THINKS THIS CANNOT BE FIXED, BUT IT IS AN ISSUE THAT WE SHOULD BE WORKING ON BECAUSE HILLARY IS THE FRONT-RUNNER AND WE WANT HER TO WIN THE GENERAL ELECTION EVEN IF WE ARE SUPPORTING ANOTHER PRIMARY CANDIDATE. That part in bold, that's the take home message. Don't put your head in the sand about this polling data because it is still early and not too late to address the situation.
Here is some additional reading on the general topic:
http://www.huffingtonpost.com/bob-burnett/good-bill-bad-hillary_b_51925.html http://www.vindy.com/content/local_regional/288904365315258.php http://thecaucus.blogs.nytimes.com/2007/06/04/a-breakdown-on-clinton-ratings/ http://thecaucus.blogs.nytimes.com/2007/05/25/2008-timescbs-news-poll-results/