Is the correct usage of “Diagnose (verb)” losing its ground?

Like many words, "diagnose" has acquired a new sense. The use of this sense has risen rapidly over the past 40 years. Here is a Google ngram graph comparing “diagnosed with” (the contested usage) with “diagnosed as” since 1970:

The objected-to usage has experienced a dramatic rise and the traditional usage has been experiencing a steady decline since 1985. It seems the new sense is rapidly eclipsing the old one.

The Merriam-Webster Dictionary of English Usage provides justification for the usage:

Lurie 1927 disapproves of using the verb diagnose with a person as its object, even though there is often no other way of avoiding a stilted sentence. Several other more recent complaints have been lodged against this use, but not all of the commentary has been negative. Evans 1962 gives his approval: “there is bitter wisdom int he popular usage, for a man and his sickness ar one.… It is a soothing fiction that the man “has” the malady; too often the malady has him.”

Because this use sounds very familiar in spite of the fact that we have only a relatively small amount of evidence for it in our files, we believe that it is more frequently found in speech than in writing. However, the usefulness of this sense of diagnose is manifest, and its use in writing may well increase.

… the women… have diagnosed themselves accurately —Joseph P. Donnelly, M.D. Redbook, March 1964

…Yang was diagnosed mentally ill—Alan M. Dershowitz, Psychology Today, February 1969

Cindy was first diagnosed a year ago… as having a tumor—Janice Eidus, Johns Hopkins Mag. May 1977

Dr. Root, the first to diagnose me correctly —Nan Robertson, N. Y. Times Mag. 19 Sept. 1982

Many more examples can be found in the Corpus of Contemporary American English, and the original questioner’s copious dictionary citations show that this usage is well-accepted even in dictionaries. The Merriam-Webster Collegiate dictionary also gives in sense 1b “to diagnose a disease or condition in <diagnosed the patient>”.

Since words acquiring new meanings happens continuously, it is curious why some new meanings garner objections while most do not. This new meaning is useful and frequent, and provides no ambiguity or other alleged deficits of some other criticized usages. Frankly, if you are going to make pedantic objections, one should eschew the verb diagnose altogether, since it is a new-fangled word that only dates to 1859 and is a back-formation (gasp!) from diagnosis. In my opinion, it should be used freely, and we can let the critics pat themselves on the back for avoiding it and feel superior to those who use it, while the rest of the world goes on using it, oblivious to the critics’ pointless objections.

My perception is that diagnose is used in the "correct" way when speaking or writing about the actions of a doctor.

When omitting the agency of the doctor, as in your first example,

The patient was diagnosed with cancer.

the alternatives are:

The patient was diagnosed as having cancer.

or

Cancer was diagnosed in the patient.

The former seems to stray even farther from the cited usage. The latter dramatically changes the subject and emphasis from the patient to the cancer.

I personally find "the patient was diagnosed" an acceptable usage, if the alternative is to always reference the doctor. That a doctor did the diagnosing can probably be assumed and would seem superfluous, especially in news reporting, where brevity is essential.