Vocal reduction effects assume that the vocal is a centre panned mono sound - in other words, that part of the audio signal that is the vocal, is identical in both left and right channels. That would normally be the case if you made a stereo recording of someone singing with a band, without applying any effects to the vocal.
The example track that you posted has a huge amount of “stereo reverb” effect applied to the vocal, so the vocal in the left channel is not the same as the vocal in the right channel. Because there is hardly any commonality between the vocal in the left channel and the vocal in the right channel, it is not possible to make automatic removal of the vocal. The vocal reduction / removal effects will fail with that audio.