Zee's QFT in a Nutshell, and Srednicki's QFT use "nonabelian."
Lancaster & Blundell's QFT for the Gifted Amateur, Tong's Lectures Noted on QFT, Schwartz's QFT and the Standard Model, and Peskin & Schroeder's Intro to QFT use "non-Abelian."
Three specific answers:
nonabelian and non-abelian are roughly the same frequency over the past few decades but noncommutative is by far the standard nowadays
nondegenerate is somewhat favored over non-degenerate
hyperkähler and hyper-Kähler both seem to be used in mathematical literature with no clear higher frequency.
That is what is, but you're probably ...
It depends on what you are trying to convey. Part of the difference comes down to whether the speaker is a crowd or several individuals, and the other part involves the distinction between direct and indirect speech.
In terms of the quotation marks, you use two different options.
The option in #1 and #2, with quotation marks at the beginning and end of ...
This is a matter of style, not grammar. To start, The Chicago Manual of Style (17th ed.) does not say to use italics with the proper names of individuals or companies (unless specifically emphasizing the name as a word or phrase, for instance).
The following is what Chicago actually says about the various points that have been discussed in comments under ...
You have a noun phrase built up of a base (cracking) and an attributive noun phrase (stress corrosion). Stress corrosion itself is a combination of the base corrosion and the attributive noun stress). Attributive nouns describe the nouns they're paired with, so stress corrosion describes the cause of the cracking (ThoughtCo).
It is not a proper noun, ...
Parliament can refer to any legislative government body. Its origin is in an assembly gathered to talk.
c. 1300, "consultation; formal conference, assembly," from Old French parlement (11c.), originally "a speaking, talk," from parler "to speak" (see parley (n.)); spelling altered c. 1400 to conform with Medieval Latin parliamentum.