Well, higher=better is simply more intuitive. There's a reason why it was adopted, and it wasn't because of change for the sake of change. It's the same reason people decried THAC0.
No no no no no no no no no no. No.
The problem with this, higher is better, is that you no longer have an upper limit. This creates problems for d20 starting around level 18 and getting worse in Epic. Further, for drama purposes, the game keeps increasing the DCs of skills for higher level characters. Yet, it's the same task. The reason they do this, though, is to keep that chance for failure.
In other words, climbing starts as a DC 10 but slowly increases with level so that there is always a chance of failure. Hopefully its a slowly decreasing chance of failure but while this works for drama, the system now doesn't work for itself because the same task is different DCs.
In Alternity, by having lower is better, you can have upper limits. Further, you get to a point where they are so good that you can see that one more skill rank doesn't help. It's the rank benefits that matter more.
Sorry but this is one of those things that I thought Alternity did very well and would not want to see changed. I am find with the dice modifiers being different or whatever but not the overall idea of it.
As to the dice part, I don't disagree it's in large part due to the D20 system. It's an elegant system, which is something I think a lot of game designers either by choice or by ignorance to the problem, ignore. However, it is a distinctive part of the Alternity system, and that's why I'm ambivalent. I will point out though that Alternity was basically the basis for the D20 system anyway. Roll a d20, add a modifier, done. The only difference is the lower=better holdover from TSR, and dice instead of static steps. As to the last part, if you believe that a d6 = 3.5*, then it's even more the same thing.
Sorry, not following some of this. Which is an elegant system? Higher equals better? I disagree. In Alternity, you say that skill checks are low but you want rolls based on the results of those checks high. That takes care of it and explains it all. Still one sentence and just as elegant.
Again, I'm not sure you are realizing the point of the modifiers. If you are, I apologize for the following explanation.
If I tell a DND player that they have to roll a d20 and add 10, without knowing the DC of the roll, that could be good or bad or anything in between. Let's say they said they wanted to attack a creature. Without knowing its AC, the player doesn't know if this is good or bad. (Assuming a new player who doesn't know the AC or understand levels completely.)
Now, think about Alternity. A player says they want to attack with a weapon they know their character isn't good with. I tell them low is better but to pick up a d20 and a d12, roll them, and ADD them together. It doesn't take a math genius to realize, for the player AND character, that this isn't a good idea. So, instead, they say they want to attack with a weapon the character is good with but have to run and get it. Sure, they have a slight penalty for running but now it's roll a d20 and d4 and add them together. Again, it doesn't take a math genius to realize this is better for the odds.
Essentially, the dice were a hint how good or bad of a chance they had. I had many players, when we were playing, look at what they had to add together and rethink what they were going to do to have better odds. Sure, the character wouldn't have that time to think but for the game it worked out well. Once in a while, they might take the shot with the +5 die penalty to see if it works but most of the time they waited or did something to reduce that penalty.
I don't have too much time to go into detail right now (stupid work) but I think my biggest perceived issue is that rolls are already swingy due to the d20, and with the modifier being swingy too, I think it creates too many situations where something logically should work (d20-d20, for example) but in practice that still leaves a large window for failure given such a hefty bonus. Like I said, it's perceived though; if someone wants to run numbers, I'd love to hear it.
In some ways, I can't argue with this. It is a big range to roll a d20. I completely agree. Further, rolling a one on a d20 in Alternity sucked when even the d4 rolled a 4 and potentially pushed an Amazing result at least to a Good and maybe an Ordinary result.
What I can say, though, is that odds for at least an ordinary result were good and that's a good thing. I can't stress that enough. Do NOT take away from Ordinary successes. They are still a success. For a while there, we wanted to see Good or better successes and I think that really pulled from the Ordinary, which is still a success.
Further, since I wanted a more heroic feel, that heroic flavor is in the ordinary successes in that they keep the battle going longer and aren't as deadly as the Good or Amazing results.
You might understand all of this about the Alternity system and if so, I apologize for the long winded answer. But I think these are important points and central to how Alternity worked and worked well.