Happiness calculation must have changed in the player's favor at some point

[anonymissimus]

I just won some mission with 1 out of 8 civilians surviving. I remember that such an outcome used to reduce nation happiness. It was 3/8 that was needed for avoiding happiness loss. The game got a good deal easier here it seems. It may be an in intended change of course. Perhaps it's also important that it was a crashed UFO mission, as they work differently (e.g. they don't reduce happiness at midnight). (Perhaps my memory stems from non-crashed UFO missions ?)
As a result, I managed to reach maximum happiness late July already. I recall that this is some months earlier than it used to be.

[anonymissimus]

Well, it is the result of commit
8679ae58a7e12f0fbf18e3776a090373c98b5a20 (2018-02-15)
The changes were originally motivated by a bug reported by me (bug 5543).

Now my problem is: No matter how many civilians died, as long as the mission is won, performanceAlien is high enough to get a happiness gain anyway because of this victoryBonusPerAlien = 0.1. While this is not really a bug, it makes the campaign a lot easier. In particular, there is no need to actively try to save civilians in difficult missions, it's enough to wait in smoke, scout every turn and kill if some alien shows up. This generally will kill most or all civilians, but it doesn't matter much anymore, see above.
I think performanceAlien should be normalized similar to performanceCivilian, which can never be lower than -0.5 even if they're all dead, so that it is insufficient to only win the mission.