Those thing might and will make a confusion for newbie.
The implication being that new users are somehow incapable of understanding the issues, and therefore it shouldn't be put out there to confuse them. Only three or four word answers please. If I were a new user I would be insulted.
He also went on to imply that the way he decides things is to get multiple opinions and try the most popular ones out for his self. This is a very poor second to educating yourself on the question and making your own evaluation with your system in mind. That is especially true in a forum that has a high percentage of new users and the question is of necessity system specific. Decision making via popularity contest is at best inefficient.
Unlike the person I was responding to, I believe new users can deal with information and answers that go beyond a yes or no, without going into a fog of confusion. Otherwise there would have been no point in me bothering to write and make the posts referenced in the first place. But if you do think new users are that dim witted, then illiterate is an appropriate term. There is a big discrepancy here. Why would you trust people like that to guide you about what OS you should be using? The logic escapes me.
As far as disagreeing with the link you posted, I don't disagree with anything in it that I remember right off hand. The point would be that they were testing only for differences caused by the 64 bit/32 bit operating mode. They intentionally didn't want other things to come into play to skew the results. My point is that system speed is affected by many factors other than word length.
It is still not common for people to have so much installed RAM that the example I gave rarely comes into play. Long documents and video are just two examples where you could easily see this situation. 30% more or less is a lot of memory use difference and could easily put you in this edge situation if you had 1 to 3 Gig installed.
I still say people should evaluate their own system and workload environment before making the switch just because others have told them 64 bit is faster. As I pointed out, it is faster on some hardware and workloads, but it certainly isn't on others.
Something else I didn't really address is that much of the first hardware that would run 64 bit CPUs still didn't have the buss width to handle a true 64 bit system. The result being that you get all of the downside of 64 bit, high memory consumption, poorly optimized software, not as much software to choose from, etc., and almost none of the advantages.
Actually, I am am a 64 bit advocate, but only when it is going to do me some good.