Whether or not we strive to go back to climate not seen for the past 50 million or even 500 million years, we might end up getting there regardless. Even if we decided today to put all our industry together, we'll still hit double the amount of warming we're experiencing today. If that is extreme enough for a catastrophic feedback loop like the collapse of the Greenland ice sheet, permafrost outgassing, catastrophic ecosystems collapse in a subcontinent, or all the plankton dying at once because the oceans got too acidic, then even our best efforts can not save us from that reality.
And once we are in that reality, those of us that survive deserve a chance. Even if Antarctica melts and 90% of all human structures ever built are flooded, even if the forests die and global dust swarms blanket the planet until even the tropics are buried in snow, even if no living thing can exist outside a purified climate-controlled space. We, here and now, owe the people living through that our best effort. And that means looking at those futures unflinchingly, determining which are more or less likely and trying to prepare for all those eventualities as best we can.
The time to throw up our hands and say "all of these options are terrible, we should just stop climate change as hard as we can" was twenty years ago. Barring a technological miracle like the Silicon Valley AI God actually saving us from perdition, a billion human deaths would be us getting off easy. We need to prepare for the inevitable catastrophe, because even if a billion humans die, the next billion matter just as much, and the billion after that, and after that, and after that and after that and after that, and after that.
(That said, obviously research shouldn't be used as a smokescreen for delaying carbon emission reduction. It's crystal clear what our politics and our economy need to do if we want to raise the average life expectancy of children born today above 50).
A good reason would be to get ahead of the bad press and control the narrative. Even something as minor as a bad turn of phrase in an internal e-mail could force them to make a press release early, in that case becaues of the risk of it being stripped of context and leaked to the press by corporate spies or well-meaning whistleblowers in a way that looks way worse than a promise to get around to it later.
Not sure how likely this is compared to it being a fig leaf over cancelling the target altogether.