I have the "alpha" version of this that Sascha did for Magix called AM-Track.
Nah, no 'alpha'. Not even close.
AM-Track is 'ok' from my today's perspective. Has become sort of classic among the Samplitude user base, perhaps for a reason, dunno. I couldn't do better back then (2003 mainly!). But we devs also wade through deep learning processes all the time, and so I wouldn't want to repeat prior work. What bothers me a lot is that it isn't oversampled, therefore it doesn't emulate a 'real' magnetic hysteresis curve driven by HF bias. But for a tracking plug that consumes almost no CPU it's still pretty good, I think. But tape emulation has come a long way in the meantime, as has research on the topic. And, computers have so much power these days that you can mimic a lot more of the tape aspects and ingredients which matter a lot in getting the 'glue', 'gel' and 'smear' and 'oomph' that is so charming about that medium, apart from simply distorting things.
That being said, I can't comment more on Satin *at this time*, as I'm in the middle of development here, among other projects, and I don't want to put too much pressure on myself. I hate doing things in a hurry, so I take my time in constantly thinking stuff over and trash ideas when they don't seem work out right, and change perspective or try differently. I'm perhaps only good at empirical programming, but happy they let me do it at u-he