you are not scraping directly...you are taking 3rd party abstracts of publisher content and using them against the publishers. this is actually even worse than scraping directly because there is no way to opt out of Mahalo "borrowing" a publisher's content without the publisher also blocking Google from indexing it.

keep in mind all of your content is creative commons licensed...so if you keep up your scrape and nofollow attribution game, you are going to be in for some surprises where you feel the effects of what you are doing onto others.

your choice...either fix the nofollow anti-attribution issue yourself or have the market fix it for you.