The usual maximum mass that is quoted for a white dwarf progenitor of $\sim 8M_\odot$ is empirically determined from young, local star clusters that have (roughly) a similar composition to the Sun. Any theoretical calculation of that maximum mass are quite dependent on the prescription of mass loss during the red giant branch and the asymptotic giant branch, which is iself highly uncertain, so the empirical limit is preferred.

Having said that, we would need some reason then why a $12 M_\odot$ star might lose more mass than similar high mass stars in local open clusters? I can think of two possible reasons. Firstly, mass loss is highly dependent on metallicity. A star with a metallicity much greater than the Sun might be expected to lose more of its envelope during the giant phases and thus could still leave a core lower than the Chandrasekhar limit. Second, binary interactions could lead to stripping of the progenitor star's envelope by a binary companion. This is the scenario thought to lead to low-mass helium white dwarfs, despite there being no evolutionary scenario for single stars that could produce such objects in the lifetime of the Galaxy. I would assume that if such interactions took place prior to the progenitor forming a C/O core, it could lead to the formation of a white dwarf.

In terms then of what environments might favour these scenarios - high metallicity environments (the Galactic bulge?) or regions featuring lots of binary stars (though most high mass stars are in binary systems).

The usual maximum mass that is quoted for a white dwarf progenitor of $\sim 8M_\odot$ is empirically determined from young, local star clusters that have (roughly) a similar composition to the Sun. Any theoretical calculation of that maximum mass is quite dependent on the prescription of mass loss during the red giant branch and the asymptotic giant branch, which is itself highly uncertain, so the empirical limit is preferred.

Having said that, we would need some reason then why a $12 M_\odot$ star might lose more mass than similar high mass stars in local open clusters? I can think of two possible reasons. Firstly, radiation-driven mass loss is highly dependent on metallicity. A star with a metallicity much greater than the Sun might be expected to lose more of its envelope during the giant phases and thus could still leave a core lower than the Chandrasekhar limit, leading to a stable white dwarf. Second, binary interactions could lead to stripping of the progenitor star's envelope by a binary companion. This is the scenario thought to lead to low-mass helium white dwarfs, despite there being no evolutionary scenario for single stars that could produce such objects in the lifetime of the Galaxy. I would assume that if such interactions took place prior to the progenitor forming a C/O core, it could lead to the formation of a white dwarf.

In terms then of what environments might favour these scenarios - high metallicity environments (the Galactic bulge?) or regions featuring lots of binary stars (though most high mass stars are in binary systems).