I'm not sure why every movie would do that,
although I have seen it happen with the occasional movie.
It seems that on occasion google will block a search or at least not return any results, this issue is beyond my control.
From the next build the scraper has a double check that will attempt to get a result from imdb if google returns a blank.
As for accuracy, I would expect about in excess of 90% of identifications to be correctly identified, probably nearer 98% using your folder naming convention.
Also, although the filename should mostly return the correct title, since the folder also contains the movie year, have you selected the option to use Folder Names when scraping in the movie preferences.