I know this subject has been discussed in great detail before. But if appears that the majority of the weight in the Algo now is a friendly SEO url with the keyword in it. Backlinks have become secondary?
I did several three word queries. Some of them had shopping intent ("brown leather widgets") and some of them were informational ("battle of widgetburg"). All of them would have significant search volume and plenty of competition.
I did see lots of urls with keywords and very few urls with question marks. It certainly didn't appear that the sites with keyword urls were relying on them alone. There were also lots of backlinks involved. And having the exact perfect url didn't seem to matter (/brown-leather-widgets). I saw far more like /department10/widgets/brown+leather/ and widgetstore.example.com/fooxxx/baryyyy/brown/browse.jsp
For the historical queries I did, I even saw some with shortened words in the urls: history.example.com/wdtsburg.html
Certainly keywords in urls can't hurt you. Even if they don't actually help with the rankings, they force you to really think about the keywords you are targeting and to make your urls simpler and more stable.
Keywords in urls also help users because they get an idea what the link is about when somebody pastes a url into a forum.
My gut tells me that Google has done enough work removing sites boosted by link schemes (penguin) that the sites that are ranking now are SEO savvy (have put keywords in urls) but more whitehat than what had been there previously.
Real simple to test this theory and to find out first hand what is reality.
Step 1 - Go to Blogger and register a site for a made up word like "goodroipow" (don't use that example since Google just crawled this page and now knows this word) and publish it with one page of content.
Step 2 - Go over to Wordpress and register a similar named site also with one page of content. Then get a few backlinks to this one, maybe even promote it in social media.
Step 3 - Sit back and monitor the results.
Step 4 - Play around with the variables and see what happens. Does it make a difference if you run a second test and reverse which site goes on Wordpress and which one is published on Blogger? Does it make a difference if the keyword is in the url but not in the page content? Get creative and keep testing.