Kemper has had great customer service since day one, if you used the proper channels. I don’t know how “personal“ it was, but any question that I ever had was answered quickly.
I actually appreciated all those years when CK himself was scarce to nonexistent on the forums, here and on TGP. There’s something cloying about the owner of a business getting in his customers’ faces, rather than allowing them the freedom to express themselves, whether it’s to share positive experiences or frustrations.
But the fact is, so many bits of misinformation about this device have been shared on such a regular basis, over so many years, and I guess it finally reached a bit of critical mass with the rabid fan base of an announced device that purported to do CK’s magic better than the original. Of course that’s at best debatable, but I can understand him becoming a bigger part of the conversation at this point.
I don't know about all that really. Christophe has been in and out of forums since the unit released, helping people, correcting misinformation, guiding people. I don't think the QC has caused any more or less misinformation or opinionation.... er... that's not a word, but you know what I mean.
I think I said it before, but in the last 10 years there has been some very impressive advances in machine learning, deep learning, and music-information-retrieval technology.
If accuracy is the goal, you would certainly look at modern techniques and tools such as the variational auto-encoder. Here's a good video that gives a nice explanation:
Here's another one:
I've been experimenting with neural networks in recent months using Tensorflow and the Python programming language.
Once you know a little bit about this stuff and when you look at technology used in software like Izotope's Neutron and Ozone.... you can start to see problems all over the place that you could address with some neural networks and a few data sets.