I've used a Tamron 150-600 mm and I didn't find it "really soft" at 600mm certainly usable that was my experience.
We can debate what I meant by the term 'really soft'. My intended meaning was more like, 'in reality' soft, or 'actually' soft, rather than 'very' or 'particularly' or 'unusably' soft.
Most reviews on the internet mention there is a fall-off in sharpness at 600mm, but tend not to dwell upon it.
Only one review I've come across shows a 100% crop of a scene taken at 500mm, interpolated and compared with a 600mm shot of the same scene from the same position, using the same copy of lens. Both shots appear to show the same amount of detail to me.
However, the reviewer didn't compare an interpolated 400mm shot, with either a 500mm shot or 600mm shot. One presumes there would be at least a marginal improvement in captured detail at 500mm or 600mm, compared with 400mm, but how significant that would be, especially when compared with a 400mm shot from the updated Nikkor AF-S 80-400 VR, which is the lens that I use, is an unanswered question.
I get the impression from most of the reviews that there is an underlying assumption that this Tamron lens at 500mm or 600mm must be better (sharper and more detailed) than the nearest equivalent zooms from Canon and Nikon, such as the EF-S 100-400, and AF-S 80-400, but no-one takes the trouble to make a direct comparison, interpolating the 400mm shots. It's a bit like assuming that because a camera has an unprecedentedly high ISO setting, the noise must be lower than in other cameras without the same high-ISO setting.
For Newbies who don't already own a good copy of a Canon 100-400 or Nikon 80-400, and who also shoot jpeg and rely upon in-camera processing, the issue is different. Cropping and interpolating a 400mm shot takes time and skill. The Tamron doesn't have to be 'better' at any particular focal length because the price is already better, or at least better than the price of the Nikkor AF-S 80-400. That may be a sufficient reason to buy the lens.