I don’t see why marking something as captive/cultivated triggers any other changes to the observation. I suspect this has been debated before, but it’s relevant to my recent experience.
Now that I’ve marked a number of observations as captive, they are also marked “casual” and thus lacking media, location, date, or some other criterion, when they have all of the above. The plants I’ve observed are totally integrated into the urban environment and interact with “wild” species. They are no less a part of the ecosystem, and the observation is no less valuable than any other.
This is related to bigger philosophical questions about what constitutes wildness. Given strong arguments that the Amazon rainforest itself was largely anthropogenic, I lean toward the position that the distinction between wild and captive is porous at best.
But there is a more practical reason not to downgrade observations of clearly captive organisms: users–particularly beginners–won’t mark them as captive if that means they won’t get identified.
At the request of a fellow user, I just batch-edited dozens of observations to mark them as captive. Now that I know those observations have been marked as defective, I can see how a less principled or less experienced user might change them all back to “wild.”
My (inexpert) preference would be to make it possible to mark observations as captive without changing anything else about them.