I'm not sure your angst is with Transcribers or Enumerators, as the first source of the issue is when the Enumerators write it on the census form. Your example of Fennellow says to me Enumerator.
You make some good points Tim, and yes I do believe much of my 'angst' can be directed at Enumerators. The other day for instance I came across the most beautiful set of handwriting when checking out an original Ancestry Census. As I also needed to go back to the previous pages, and even forwards, it was a pleasure to see such clear and distinct handwriting. My wife - as I have mentioned before in the Forum - writes copperplate (I think people enjoy her C card envelopes more than the cards) and I called her to view the writing. She said something to the effect, that that is how people should write. But sadly, it is a rarity, and some handwriting (and probably my own) would task the best Transcriber. However, that said, and - as you also mention - all to often the surname is obvious to me and with the best will in the world should be to everyone else; but it seems this is not the case.
I don't know what skills are required of an Enumerator, but hand in hand with good education and spelling skills (knowledge of the area of course) should be the ability to provide clear handwriting, making transcription easy.
Years ago -but not as far back as the Censuses available to view - the Computer age dawned and the first manifestation of same were machines that could read punched cards. These were of course produced by operators interpreting handwritten script, or if they were lucky typewritten. If the punched card operator wrongly interpreted something, that could be a mistake in perpetuity. So it was quickly learned that cards needed to be verified; requiring a different operator to go over the same script. By this method mistakes were picked up, although not removed entirely as a second operator could make the same mistake as the first. But but by and large errors were few and far between.
Sadly time and money is not available to verify transcripts, so in effect we Researchers do the job for them. So when we judge a transcription to be anything from a pure abomination to an unlucky mis-read, we reserve the right to voice an opinion.