In the U.S., the inventor is required to submit citations for all
relevant prior art that the inventor knows of. It is in the inventor's
interest to aggressively seek out relevant prior art and carefully
distinguish, in the specifications, the differences between the prior art
and the claimed invention. This is more true for prior art that is close
to the invention for, if that prior art is cited on the front page, it
cannot be cited by an opponent in invalidation proceedings.
Now, in my limited case, my applications were rejected on the first pass,
but granted after rebuttal. Corporate patent attorney's I have talked to
view this scenario as good because it gets more prior art on the front
page. Since there were rejections in all three patent jurisdictions, I
would argue that all three systems operate in more than a simple
registration mode. I did find variation in quality in the examiners'
citations. The European examiner's citations were the most relevant and
essentially covered the same span of prior art that I had uncovered in my
research - same authors, same work, different publications. The U.S.
examiner's citations were further afield and suggested a difficulty in
understanding the underlying technology. The Japanese examiner's
citations were more narrowly focused and were of applications that had
been submitted but not published when I filed - essentially interference
applications.
Now to your argument that the lack of non-patent citations indicates a
lack of serious searching. This really has no meaning. In many areas,
R&D that is going on in companies is well ahead of that in academia. For
example, when I was in Dupont working in the Kevlar group, related
academic work was 10-15 years behind. In other areas, for example,
automotive repair putties, there is no significant academic interest, nor
should there be.