If there are interactions, then the ecologies start merging and adapting. The native species can still thrive because there is participation among both old and new.
Seriously, go look at the iceplant covering California's coasts [1]. Or the himalayan blackberry all over the pacific north west [2]. Or the kudzu vine all over the south east US [3]. They smother native competitors and eliminate them everywhere they go, turning once diverse ecosystems into monocultures.
The ecosystems don't adapt, they get wiped clean of any native plants whose niche overlaps even slightly with the invader. If the native species could adapt, the introduced species wouldn't be called "invasive" but "naturalized"
[1] https://www.fiscaliniranchpreserve.org/images/Photo%201%20NP...
[2] https://extension.oregonstate.edu/sites/default/files/styles...
[3] https://upload.wikimedia.org/wikipedia/commons/4/4c/Kudzu_on...
A typical knotweed patch that I see here in Scotland is in a disturbed urban area, for example by the railway. While it might be displacing a native species that takes over ruderal sites, e.g. blackberry or gorse, it still performs its functions of urban greenery - makes habitat for birds, cleans up the city air a bit, covers up trash on the ground.
Some of the roots I've dug out had about 8cm diameter and they were tough stuff.
No you don't want this stuff even in urban wasteland.
Or growing in railroad ballast where it's can cause a very expensive clean up operation or destabilize the foundation of the railroad.
https://en.m.wikipedia.org/wiki/Pontederia_crassipes
Actually the article says it is invasive in many countries.
You have to introduce species that bridges the gap and creates those links.
To give an example with kadzu:
https://www.smithsonianmag.com/science-nature/true-story-kud...
Tldr: kadzu turns out not to be as invasive as people thought, and the kadzu bug has been eating its way through them.