After what Globe Collector said in some post of pinked and redded out fluorescents and CFLs, I'm confused, and don't know if all pinked out lamps are air poisoned lamps, or some can really deplated of mercury, and if there is such a thing "Mercury starved" fluorescent lamp.
Also, in the "Fluorescent lamp" article in Wikipedia , one of the EOL mechanisms that appears, is loss of mercury (Meaning mercury starving), and I will quote:

"Loss of mercury

As in all mercury-based gas-filled tubes, mercury is slowly adsorbed into glass, phosphor, and tube electrodes throughout the lamp life, where it can no longer function. Newer lamps now have just enough mercury to last the expected life of the lamp. Loss of mercury will take over from failure of the phosphor in some lamps. The failure symptoms are similar, except loss of mercury initially causes an extended run-up time to full light output, and finally causes the lamp to glow a dim pink when the mercury runs out and the argon base gas takes over as the primary discharge."

Either of the EOL mechanisms would become the killer one, if they had the chance. It depends, which "strikes" as the first one.
Historically the weakest part were the electrodes: Making seals properly is a matter of sufficient care for the process and the mercury losses were easy to overcome by just dosing really an excess of it. So the electrodes remained as the weakest point.

But the electrode design progressed a lot, mainly with the ability to extensively study and simulate the sputtering in details (so they could be engineered to be way more robust without too much sacrificing the manufacturing cost) and their life was prolonged by the use of programmed start ballasts (when the lamp has to sustain high number of switching losses). So as the electrode life was a focus for decades, with top products it become comparable with the installation life ("XXT" from Osram and similar 80+khour life could last for 10 years in an office installation, the fixture life is similar).

But the environmental pressure and the need to operate the lamps at higher temperatures makes the excessive mercury dose rather problematic. And here popped out the problem of the mercury loss, which was for decades just neglected (because adding an excess of the mercury was that simple). And with the market pressure coming from the LED's, I'm not convinced there would be any significant motivation to invest into that area.

And the air leak is a real manufacturing quality problem: It is to big extend caused by the pressure for productivity of the manufacturing line, when there is no real upgrade to the machinery (the LED pressure means the return of investment would be very questionable). So the process people just "cut corners" in points restricting the machine throughput and that are usually the glass stress relieving gradual cooling and glass forming steps: Instead of slow gradual heat up/cool downs, these get "hurried" and the material stresses and thin bubble seals are the result.
I would guess these sealing faults would become on par with the mercury loss, as the mercury loss won't be overcome and there would vbe still strong pressure to keep the mercury dose low, so the glass processing would be made as fast,till the sealing problems would be just comparable to the mercury loss.

I might be completely wrong but aren't the first seconds of a cold MV (MB) start also a sort of mercury starvation? The mercury covers the glass tube at that time.
I know that its pink colour is due to the coating but what would be a colour of a mercury starved fluorescent if the phosphor was scraped out?

I might be completely wrong but aren't the first seconds of a cold MV (MB) start also a sort of mercury starvation? The mercury covers the glass tube at that time.
I know that its pink colour is due to the coating but what would be a colour of a mercury starved fluorescent if the phosphor was scraped out?

Not at all.
The pink color of the MV come from different reason than the starved fluorescents: In fluorescents the phosphor respond with the full spectrum on just the short wave UV. But that get blocked in the MV by the quartz, outer bulb atmosphere and mainly selfabsorbtion in the mercury.
With the MV the phosphor is actually designed to glow pink, because unlike the fluorescent, the (fully warmed up) MV gives out high power in the visible MV lines, which need the pink glow from the phosphor to complete the desired white light. And as the mercury at lower pressure does not emit as much of the visible lines, starting up lamp appear pink.

And even when accepting the mind game and judging what happens there ridiculously according to what should happen in a low pressure fluorescent. And even regardless, how you define the "starved":
"Not enough mercury for normal operation": It means after warmup, so when stable. As it is just starting to warm up, with this definition it could not be starved.
"not enough Hg for a low pressure discharge": With that definition it would be rather "argon poisoned" (too much Ar for a proper low pressure Hg discharge). And heavily "overloaded", so the mercury plasma get selfabsorbing for the UV.