Disagreement is not inherently antagonistic nor condescending, and nothing about my statements was intended to be, nor could it be reasonably interpreted as such.  There is nothing belittling about expressing a difference in taste, especially when it is clearly stated to be only a matter of preference.  I hope you're not the kind of person who takes offense at any differences of opinion, but if you are, I'm not going to waste any effort catering to you.

You didn't start this thread, so please don't tell others what the point of it is.  From the title of the thread, and from the original post, the subject is much more broad than simply a mathematical breakdown for purposes of optimal power-gaming.  Different people have different preferences and different play-styles, and those differences are bound to show in a thread like this, and that's fine as long as nobody tries to insist that the thread focus solely on a particular set of preferences.

In addition to the originally stated purpose of this thread, I see another important purpose: this thread could highlight areas of the game which lack balance: if everyone is selecting or eschewing particular skills or stats (such as taking lots of endurance early and then neglecting it), that is an indication of a balance problem which needs to be corrected in order to make alternate character-building choices more attractive.  Ideally, codezombie will watch this thread (and others like it) and tweak things to undermine any optimization strategies which are widely agreed upon, until there is no longer significant consensus about optimization strategies.  There shouldn't be such a thing as the best way to build a character.

As for agility, if you truly want to conduct a "by the numbers" analysis, then you lack sufficient information to judge the utility of points spent on agility.  You've made qualitative assessments (which don't refute my points), but you don't really know what the mathematical impact of agility is.  I will demonstrate what I'm talking about with an example, with a deliberately simple scenario:

You are standing at one end of a corridor that is 1 square wide and s squares long, with an enemy at the other end which is only capable of melee attacks.  Since the corridor is s squares long, that's how many steps it will have to take to reach you.  During that time, you will be able to attack the enemy with your spell(s) n times, where n is a function of your agility.  On average, each attack will do d damage, where d is a function of your magic stat.  So, the average amount of damage you can deal to the enemy before it reaches you is t=nxd.  If n>d, then t will be increased more by raising d by a given amount than it would be by raising n by that same amount, and vice-versa.  Right now, we know the formula for d, but we don't know the formula for n, so it's impossible for us to say with any degree of certainty the relative impact of raising agility or magic at any given point in a character's progression.

Even if the enemy is much faster than you, the same math applies, although it does get considerably more complicated.  If the enemy is fast enough, n will be equal to one no matter how fast you are, or maybe even zero.  Your speed still matters, though: once the enemy reaches you, it will start attacking you, so then the question becomes: "How much damage will this enemy do to me before I kill it?"  This also applies in the scenario of an enemy who can attack you from a distance.  In either case, for every attack made by the enemy, you get to make m attacks; if the enemy is faster than you, then m will be a fraction.  If the enemy has e hit points, you will have to attack it a=e/d times on average to kill it.  During that time, it will attack you a/m times.  On average, each attack it makes will deal h damage, where h is a function of your armor, which means it's a function of your agility (or if it is a magical attack, it's a function of your magic resistance, which means it's a function of your magic and endurance).  On average, the enemy will deal hx(a/m) damage to you before you kill it, which expands out to (hxe)/(dxm).  We don't know the formulas for h or m, but we do know that both of them are functions of agility for physical attacks, while for magical attacks one of them is a function of agility and the other is a function of magic and endurance.  We also know that the most effective way to increase the denominator is to increase whichever is smaller out of d and m.  If the enemy is much faster than you, then m will be very small, indicating that adding agility is likely to give you more bang for your buck, although we can't be sure of this without knowing the formula for m.  Sure, at some point the monsters will be so fast that no amount of agility will allow you to "keep up" with them, but there can still be enormous benefits to mitigating your speed disadvantage, if even by a little bit.

I think this is the first heated discussion about game mechanics I've seen.  It is good to have a variety of opinion on the game mechanics.

My goal with various skills is to make it difficult for a person to decide which ones to take, because they all have some sort of benefit depending on the person's play style.  There are a couple skills that I see a need to enhance, like the area of effect spells and the melee and close range skills.  Also, the potion enhancement skills likely need an improvement.  I have a feeling most people don't use them.

I plan to add some more potions.  I may classify them as health or life based potions and mana or magic based potions.  That way, the skills will maintain the same effects, but also improve other potions of their type.