At VoxEU, Paolo Manasse asks the following about economics blogging:

Why do many economists, especially in the US, devote a lot of time and effort to manage a blog (notable examples are Steve Levitt, Paul Krugman, Brad De Long, Greg Mankiw, Dani Rodrik, Becker and Posner, Mark Thoma, John Taylor)? Perhaps the professors, at a certain age, are just fed up with the long lags required to publish in scientific journals? Or do they simply wish to gain greater visibility, for themselves and their scientific work? Do they do it for ‘civic duty’, in order to popularise their ideas, to generate a debate and get feedback from readers? Even more puzzling, why in many European countries, namely Italy, does this not happen?

Manasse reports, based on a CEPR paper (boldface mine):

The results are quite interesting. The link from a blog increases significantly the paper’s downloads and abstract views in the month of the blog publication and, to a lesser extent, in the following month. Figure 1 shows that some blogs have a very large visibility ‘multiplier’ effect. A quote from Paul Krugman’s blog, or Marginal Revolution, or Freakonomics leads to an increase in the number of abstract views by between 300 and 470 units (compared to a monthly average of 10.3 views for NBER papers). Moreover the number of downloads increases by between 33-100 units (compared to the monthly average of 4.2 for the average NBER paper)….

For the second hypothesis, the authors use the results of a survey about the most admired economists in the US, and intersect the results with the rankings (RePEc) of the top 500 economists in the world, based on scientific publications. The authors test if the probability of appearing in the list of the most admired economists, controlling for scientific ranking, is influenced by being a blogger or not. It is. Bloggers are about 40% more likely to appear in the list of most admired economists – an effect equivalent to that of being among the 50 top world economists based on publications records.

(there’s also an argument about how blogging influences readers but I think the data are pretty weak for that)

I’ve never seen this quantified for the physical sciences (although I don’t keep track of these things–if the work exists out there, feel free to link to it in the comments). But this does seem to match my general (although admittedly biased) impression: scientists who blog are more likely to be noticed and have their papers read. Having an online presence isn’t a bad thing; it actually increases your profile. It reminds me that Steven J. Gould had far more influence on scientists due to his weekly column in an NYC-based science magazine (Natural History) and his popular books than his published papers.