It comes down to generations basically. Some people have different opinions of what counts but in a basic sense, I think we can all agree that only future generations should be counted, provided they were not assisted.
Here is a scale, from 0 (captive) to 10 (wild). Most people would count plants as wild at a score of at least 3. Some may count as low as 1. Some may not count until higher values.
0 - Original tree (planted)
1 - Shadowed offspring (growing under parent). Usually die prematurely. Generation 1 plants.
2 - Saplings from original tree, still within close sight of parents. Generally die prematurely. Generation 1 plants.
3 - Saplings from the above step maturing, and then producing more saplings. Longer-term survival. Generation 2+ plants.
4 - Generation 2+ saplings spreading into wild areas. Reproduction evidenced.
5 - Generation 2+ trees and saplings with obvious “infestation range”, spreading across into multiple sites or along a creek.
6 - Later generation saplings appearing a mile or more far from parent site, but still traceable to the original parent plants (i.e. generation <= 2).
7 - Scattered colonies across multiple sites. Spread between local areas evidenced, with self-sustaining populations probable.
8 - Several colonies persisting for multiple consecutive years. Several satellite colonies continuing to spread new plants.
9 - Several colonies adapting to various habitats, even outside of urban landscapes. May or may not be localized, but no doubt that the plant is spreading and sustaining.
10 - Widespread weed.
Technically? I think we can all agree anything above score=1 is confirmed wild and growing on its own. But what score counts for you, personally, as “iNat wild”, and worthy as a data point? That’s for you to decide, since no one shares the same opinion.