We have written an essay titled like this post that has been accepted for publication on The Information Society journal:
Research evaluation, which is an increasingly pressing issue, invariably relies on citation counts. In this contribution we highlight two concerns that the research community needs to pay attention to. One, in the world of search engine facilitated research, factors such as ease of web discovery, ease of access, and content relevance rather than quality influence what gets read and cited. Two, research evaluation based on citation counts works against many types of high-quality works. We will also elaborate on the implications of these points by examining a recent nation-wide evaluation of researchers performed in Italy. We focus on our discipline (computer science), but we believe that our observations have relevance for a broad audience.
The essay was accepted a few days ago but it should come to press very soon, probably the next month. Indeed, during the reviewing process one of the reviewers stated "it is bang on target with respect to the current situation in Italy".
We published a very preliminary version of this work on the public domain: http://arxiv.org/abs/1308.1946. Some of the data that we used as a basis for our claims are available on this site, in the data and tools section.
...just in case you were interested in "ranking" this journal: class 2 of 4 (1 being the best) for the Italian evaluation system (so-called VQR by ANVUR).