Thanks for the responses, guys. I figured this would make a decent discussion point.
@ ister:
Re: constraints on value and subjective application. I agree, it can be a problem. I haven't constrained the value of a character in terms of a number or a range of numbers. Simply speaking, some characters will have a greater range or just be more effective than others. For example, there is no question that a dual or multiclass thief/mage will become more versatile, valuable, and powerful than a straight thief. As such, I find it difficult to put a maximum overall value on a character. Even with the same character, said character may be more valuable or take on different roles in different parties. A fighter/cleric will have more value in a party of mages than it would in a party with surplus melee and healing capabilities.
As for the difference in subjective application of numbers by different people, it doesn't really matter to me what values people assign, as long as they apply the values consistently between their own characters and parties. This party valuation method was primarily designed for, and is most valuable for, roughly figuring out what the strengths and weaknesses of a party are. Moreover, the valuation of functions helps one to discern how the party will stack up to various challenges and if it may need to be changed to fit how the player wishes to play (eg. melee-heavy, stealth, spellcasters, etc). Further, the assigning of roles to party members is something we do all the time, and this numerical analysis is simply another tool to rationalize party creation.
However, in the example you gave, I would never give a 1.0/1.0 to a paladin in melee/ranged. The valuation IS rather subjective, but the main criteria I use to determine what valuation a character receives are (in no particular order):
1) Potential to be effective in the function - obviously, a paladin will never be as good of a thief as a thief, because of the nature of the class and its class-specific abilities (or inabilities). Likewise, a thief will never be as good with a melee or ranged weapon, because of Thac0 progression, weapon specialization, and (possibly) strength bonuses. Assuming relatively equal levels, of course.
2) How I (or whatever other player) intend to use the character - there is some fluctuation in this as the game progresses, of course, and on a micro level, could change between every fight. It may also be different for the same character in different parties. A paladin with a longbow might get a 1.0 for ranged weapons in a party where no other character has the ability to use longbows, for example. Or, as is the case in my current party, the shield-bearing paladin is the primary melee option who never uses a longbow because the player is too lazy to switch back and forth in the inventory screen between shield and bow all the time. If I want my cleric to use all of its 5th level spell slots on Raise Dead instead of Flame Strike, that will reduce the cleric's value as an offensive spellcaster in favour of healing abilities (generally categorizing the raising-resurrection spells as healing).
So essentially, the numbers one assigns come down to capability and use. This is a rough method of estimation, but it is surprisingly effective at figuring out the strength of a party's various functions, at least relative to the other functions of the party. Try it and see on one of your existing parties.
@ NobleNick
I should state that I have not complicated this analysis by making determinations of power, except to provide the warning about dual classes and various stages of the game. I figure that as long as the ingredients (eg. sufficient party capacity in a given function) are there, the power will be there as well.
That comes down to decisions of positional and magical tactics, which are situation-specific and don't really come down to numbers.
I will admit that capacity is generally more important than the amount of time the character performs a certain function, but it makes sense to take into account the amount of time a character performs a function, or alternately, how vital that function is when the character needs to perform it. A character specialized in longbows is useless for ranged purposes if it never uses that capability. Therefore, there has to be some consideration of how often, or in what situations a character will perform a function. Perhaps it is better to look at it in terms of importance, rather than time. A Heal spell, well-timed Fireball, or well-placed Web may be the most significant action that a character performs in a given fight, even if the fight lasts significantly longer than the one round it takes to cast that spell. In that sense, it makes more sense for the cleric or mage to get a 1.0 in healing or offensive spellcasting than to get a 1.0 in the piddly sling it uses the rest of the time.
Bear in mind that this categorization is and never was intended to be a perfect numerical model of how a party will function. It is merely a device to provide a (somewhat) rationalized method of determining how well a party may perform at certain things.
One of the ways to get around the 'inaccuracies' of a number assigned to a character would be to assign a "more specific" number, e.g., numbers in increments of 0.1 instead of 0.25. However, the pitfall in that is that it still will not be accurate, as the function of a character is more fluid, often changing from fight to fight. My Power Party currently in TotL has a bard and five warriors - given the short duration of protective spells, the amount of damage the party takes and the slow rate of healing by War Chant, I generally switch around my warriors so that they all take turns tanking. The rest cast spells or shoot missiles, depending on the situation and their own capacities. Therefore, it makes no sense to "pigeonhole" any member of the party (with the possible exception of the bard and the paladin) so closely. Striving for numerical accuracy and striving to make a party conform to that numerical designation will ultimately limit, consciously or unconsciously, the ability of one's characters to take on different roles according to necessity or whimsy.
This is by no means a foolproof method, and innovative (or patient) characters can and will get around the limitations of any given party - for example the lack of thief skills in Dragon's Eye, as in Soothsayer's recent party, or the lack of dedicated melee skills, as in my (currently mothballed) priest-heavy party. Good strategy is more important than good numbers.
I'm still welcoming any modifications, proposals for refinement, or general comments.