How much no-follow effects rankings in comparison to dofolow, is practically imposable to test.
If they effect rankings IS easy to test, just not to a defined metric.
Example:
a few weeks ago in response to someone questioning the value of mass comments, i hit a site they choose (not theirs) with 100k trackbacks. 95% are no-follow.
The site is 8 years old, ecoomerce, and in a very competetive, mainstream niche.
The target term has a traffic travis rating of 'difficult' and serpIQ score of around 60.
Before the blast, they ranked 19th.
The day of the blast, they dropped to 54th.
The next day they were up to 9th
It then danced for a few days, and has settled at 11th for the last 2 weeks.
So no-follow links, even low level comment spam, sure do work.
Just dont try this on your 2 week old autoblog!