iNaturalist shows A. floribunda in a few locations north and south Tasmania.
Also A. longifolia subsp. sophorae is very prevalent in coastal areas. I've seen it mentioned a few times as a favorable candidate based on one report but lacks thorough investigation as to whether it is a viable source (and the presence of a histamine compound may be an issue). The non variated longifolia is reported as too variable, and feeble results at best. Sophorae is very easy to identify. Though it does look quite similar to blackwood (A. melanoxylon) out of flower, both have sort of ovular wavy phyllodes and curly seed pods. The main difference is sophorae doesn't grow into a big tree and the flowers are elongated, plus sophorae can handle dry sandy environments.
I'm down south. As _Trip_ mentioned the user Rhyzobium on shaman_australis sells courtii seeds from a plant grown down here. He has other growing well outlines in this video, what a legend. (
)
We should all work on growing these species but I also want to find an abundant wild source, and Sophorae is a good candidate I think. As for A. retinodes (silver wattle), also prevent here, also inconsistent analytics, I am leaving that as plan b for now. Let's crack this. There must be a local source that can be utilised.
Currently I have some sophorae galls that were harvested in autumn, after an extended drought which is when content should peak, but I haven't gotten around to investigating that. Reading that galls concentrate tannin I thought that might also concentrate other things, however I suspect I might have been a wrong assumption and should have just collected phyllodes/twigs instead. Hopefully I'll have some answers soon.
It is probably the worst time to collect material right now as it's so wet and flowering season, the plant is diluted and focusing energy on reproduction, so will probably wait till late summer before looking for any more material. With these tentative sources I think timing and finding good specimens will be the key.
Ideally we want a source with a decent content (~/>0.5%), reliability and repeatability. Anything less wont be worth it for most people.