Well yeah it's subjective...but if I recall correctly, you were the one who introduced the term -- I'm trying to work within the assumption set you started. "Worse" implies the existence of a spectrum of bad to good; if you are comparing two points on this spectrum, the one closer to the bad side is "worse" by definition. So if you imagine non-human animals represented by a point on the spectrum and humans represented by another point, the statement "humans are worse than animals" means that the human point is closer to the bad side than the non-human animal point."Worse" is a subjective term, so what's worse to you might be better to me. Same with "goodness". How can there be a "total goodness" anyway? This does not compute.
I interpret your use of "in general" as a representation of humans that is not just one point on the spectrum, but a point for each human (so around 7 billion points). Maybe some of these points are closer to the bad side than non-human animals and some are closer to the good side. But it still seems like your statement means that when you generalize the scatter plot with a best-fit point, that point is closer to the bad side than the non-human animal point.
How else would you define a group being worse than another group? We're already pinned to a subjective framework when we start using the word "worse". It doesn't make the framework any more subjective to use the word "better" or even the term "total goodness". Thing x is worse than thing y when thing x's total goodness minus its total badness is less than thing y's total goodness minus its total badness.