martinibuster - 8:41 am on Dec 1, 2011 (gmt 0)
I agree brinked, those results say much, though less about motivation, and much about the algo. I will explain.
The OP's example is a good demonstration of how this isn't the same old algo. What's throwing off the query is the phrase "sunday hours" in conjunction with dematologist. I am guessing there is not a satisfactory answer for the query when the word "dermatologist" is in it. So Google is throwing out the word dermatologist.
In cases where a logical answer is missing Google tends to remove part of a query in order to guess at user intent. If X doesn't exist in an X, Y, Z query, Google is then guessing results for Y and Z will satisfy, since a query with X doesn't have a satisfactory answer. A better answer would be to tell you there are no dermatologists open on Sunday. But there may be no website to scrape to give that answer.
Q: Is there a dermatologist open on Sunday in Jupiter, Fl?
The above is the ideal answer. But Google isn't capable of replying no. So it's trying harder to answer the question since it can't simply say no. It probably doesn't help Google answer your question that the number one ranked site for "dermatologist jupiter fl" doesn't even post business hours on their website (indeed, many of the Jupiter, FL dermatologist sites might not even post hours on their sites- which might be an opening someone may want to plug. :)).
In the absence of logical results for your query, Google is removing the "dermatologist" part of your query. Unfortunately, that word is not a descriptive word, like an adjective, it's essential to the query. What Google is doing is blending in results for Sunday Hours 33458 [google.com], and hoping those will make sense for you, since the original query does not have a satisfactory answer. Sunday Hours 33458 is a query that Taco Bell ranks high for, which is why you are seeing it there.
That's not an example of the algo being dated but an example of the algo thinking outside of the links and content paradigm and second guessing user intent to provide an alternate set of results that perhaps under different circumstances gives what Google feels is better results.
I see this in image search when I search for specific colors and patterns in a specific fishing lure brand and model. The top results are usually what I'm looking for but they are sometimes mixed in with partial matches, the right lure but wrong pattern or wrong color. Or the wrong lure but right pattern and color. That's the algo second guessing what I am looking for. Sometimes works, sometimes doesn't.
In the OP's example, Google is second-guessing user intent, thinking outside of the links and content bun. Unfortunately this results in Taco Bell showing up for a dermatologist query. It's funny but also a good example for tearing down the algo to illustrate what it is doing.