the number of ridiculous strawmen, arguments by authority, arguments by pointing to something long and rambling that has nothing to do with the initial argument, etc. is nothing short of overwhelming.

Some things to consider on these points (mostly because I have not noticed a prevalence of these issues)

• Strawmen. If, at point X, Y looks like a strawman of a position, then at point Y, X will look like a strawman. I think. If that's the case, it could be that many of us are at point X (LW rationality techniques, etc) and you are at point Y - making valid, credible arguments that we are countering with strawmen, as it were.
• Arguments by authority. A hallmark of LessWrong is linking back to the sequences or to other posts; this could very easily look like we are saying "Eliezer said that's not the case". We aren't; he just produced a very good explanation of why it isn't the case, and it's easier to link to that explanation rather than fumble through our own duplication. Another point is that the average LWer is far more capable of deferring to people they know to be often correct - their judgement as a Bayesian reasoner is itself evidence. This looks even more like argument from authority, but there are subtle differences.
• Links to long, rambling segues that aren't related. They are related, mostly. A combination of decompartmentalised thinking, skill with readily drawing analogies, and skill with (very) long inferential distances can produce relationships that seem bizarre or unlikely.

Lastly, this comment:

it's normally to the purpose of retaining your views, not discovering where you might have gone wrong or where you might do better.

is definitely a concern for ALL LWers. I suspect you have stumbled onto a case analogous to theism: it is not the case that we wish to retain our atheism and therefore we argue to keep that view - we really, truly, have considered all the evidence and all the arguments, and we reject it on those grounds.