Why do people care about feat taxes?

[Willie the Duck]

Get ready for the usual "only PrCs in CW and CAr lose their features when they lose reqs" and "what about the Dragon Disciple" arguments.

Prescient.

1) I'd argue that even for theorycrafting, folks generally have some limits about how "pure RAW" they will be in their build -- if there are no limits other than "but the rules," then it's pretty much straight to Pun-Pun (or your other favorite infinite combo).
…
If nothing else, while RAW discussions have their own value, it's not surprising if someone points out, e.g. "by RAW, monks aren't proficient with unarmed strike," that at least some folks who are more focused on actual gameplay may tend to respond with *derisive snort*.

This thread is doing a good job of explaining why RAW-arguing has gotten such a bad name and why future editions have deliberately hewed to a 'rulings over rules' mentality.

Regardless, on some level, even theory-craft discussions have some level of 'will others accept this?' in their makeup. Usually people are putting forth a build for consideration that others might look at their proposed build (or whatever) and deem it a good suggestion. If the build is predicated on too many things that you have to argue the validity of (be it 'is this RAW?' or 'would the PCs have access to this, at any real table?'), then no one else on the theory-craft thread is likely to look at your build and consider it one of the top suggestions.

i thought it was a well established fact that 3.5 secretly trains you to be a potent lawyer?

i'm not sure this was the Rai behind 3.5 but it certainly is Raw.

I'm not a lawyer, but I work with them. It has been explained to me that that's really not what makes a good lawyer. Thing is, this 'my arguments are irrefutable!' mentality that always seems to pop up in 3e RAW debates isn't how you get good at high school debate team, much less make a good IRL lawyer. In law, you actually have to convince others of your position. These things are, really, more like the people studying computer science that Mordaedil mentioned (except with computer science courses, in the end there is a compiler that fails to produce a given output if you've coded wrong, so you actually in-the-end know if your logic was correct or not).