VST support has always been a problem for Audacity due to licensing issues. (Audacity’s CPL v2 license requires that the full source code can be distributed, but Steinberg’s VST2 license prohibits source code distribution.) Audacity has tried to work around this problem by using open source “VST compatible” libraries, but basically they’re not as good as the real thing. (OcenAudio does not have this problem because it is not open source, but the downside is that it will never be part of the main Linux repositories because it is not open source).
It “may” be possible for future versions of Audacity to resolve the VST problem, as the newer VST3 license is compatible with the newer GPL v3 license. Unfortunately, upgrading Audacity’s license from GPL v2 to v3 is not straightforward, but the VST issue provides a strong incentive to jump through the necessary legal hoops to do so.
Historically, most Audacity developers have worked primarily on Windows or macOS. Fortunately Audacity now appears to be gaining more Linux expertise among new developers, so I’m hopeful that LV2 support will continue to improve.
I’m a big fan of Jack Audio System. It has no problem matching other high performance / low latency audio systems such as ASIO / CoreAudio, and also has fantastic routing capabilities (better than ReWire), along with some other cool features that are not currently used by Audacity, such as Jack Transport and netJACK.
The main downside of Jack is that it can be difficult (or even impossible) to set up with some audio devices.
Jack always runs at ONE sample rate. To support sample accuracy, the audio stream cannot resample on the fly, so it is not possible to run two apps that have different sample rates at the same time. To change the sample rate, you have to stop Jack, reconfigure it, then restart (and hope that your hardware works with the new sample rate).
Jack does not easily support more than one audio device at a time.
Jack has an option for “realtime scheduling”, which maximises performance on systems that fully support it, but causes serious problems on systems that don’t.
PulseAudio addresses all of these problems / limitations. It can do on-the-fly resampling, support multiple devices simultaneously, has excellent support for a massive range of hardware (including OSS only devices), and is generally very robust.
PulseAudio and Jack are designed for different things. PulseAudio is designed as a general purpose audio system, whereas Jack is “fundamentally designed to be a component in a pro-audio/music creation environment”.
I have little experience with Carla, so I don’t think I’m qualified to comment.
Those are all Debian based distributions.
Audacity is distributed through official repositories by many Linux distributions, including Debian / Ubuntu based distros, RedHat / Fedora based distros, Suse / OpenSuse, Arch / Manjaro based distros, … The recommended way to install apps on Linux is to use the package manager to install the official repository version of the software. The downside of using the official repository versions of software is that they tend to lag behind the latest release versions, so if you really need the latest and greatest then you have to use some other method, such as Snap / Flatpack / AppImage / building from source. Unfortunately most of the Snap / Flatpack versions available so far have had problems, sometimes serious problems.
The Audacity developers are also working on an official “AppImage” version. It’s not yet perfect (FFmpeg support, and Help links don’t work for me), but it is already mostly working.