Good stuff! Customer perception should dictate everything. Walter Shewhart wrote a paper in which he (among other things) introduced the concept of operational definitions. In that same paper, he articulated a definition of quality that I have found to be very useful over the years. He basically said that any given thing has three types of quality:

Type 1:That quality which characterizes a thing itself independent of all other things and of human volition and interest.

Type 2:That quality which characterizes a thing A in its relation to another thing B as a part of a whole and independent of human volition and interest.

Type 3: That quality which makes a thing wantable by some one or more persons.

In manufacturing, we can control types 1 and 2; we have to rely on marketing and design to get customers to give us clues as to type 3, then translate them into types 1 and 2. That translation is what we should be about.

I'm not sure I agree that quality always costs more...in fact, I'm sure I disagree. This was the problem that US Auto manufacturers didn't appreciate in the '70s...they thought that cheating had to be the only way the Japanese could build quality at the prices they were charging. 

Quality is just like a restaurant menu: the customer can only get what the chef has prepared. Certainly the customer can choose, but his choice will never be unlimited. If I go to a chinese restaurant I expect soyabean spaghetti, but if I expect original mozzarella cheese, I'm a bit of a fool. What I mean is that no customer, whoever he is, will ever dictate the "quality" of the product or service he's going to buy: this is the lark mirror that blinds us since decades. The industrial truth is: "Your car, dear customer? Any color, provided it's black". Thank you.

Quality is: Conformance to specifications established to meet the needs of the customer or end user.  Quality may also include the reduction of variation within a range or around a nominal value if such reduction adds value for the customer or end user.

The customer drives quality

A customer may not know what their requirements are

A specification must be established

Variation reduction may or may not be important

The customer may be the final purchaser or they may be the next operator in the process

The article isn't arguing against quality being determined by the customer.  What I hear is the reminder that even  - no, make that ESPECIALLY when we set our own standards of performance, we should take a more reality-based approach.  Or, as I often coach people, "don't do A+ work on a Pass/Fail course."  We frequently spend more resources than we need to, including the most finite resource of our own time, making things better than they need to be for a given use or situation.