I already said this twice in this thread, but the problems I had with the whole "Spelunky isn't a roguelike" thing are:
1. I never advertised Spelunky as a roguelike, or even a platformer with roguelike elements, really - I said it was inspired by roguelikes.
I'd say that point is sustained, however it's not so much what you've said about it but what everyone else has said about it. Practically everyone mislabels it as a "roguelike".
2. Glenn Wichman, the creator of Rogue, has stated that "the quintessential feature of a Roguelike" is that "the adventure has to be different every time, and the game has to be capable of surprising even its creators". Not the lack of a save feature.
Well, it seems the original author is a bit deluded, since if he was surprised (positively) from its output then he did not expect it (to be that good) -- i.e., he did not design it well. Or he's too dumb to understand the possible outputs of his program. If however, understands the algorithm he's using so well and generates awesome levels every time, there would be no surprise whatsoever.
Besides that, I don't think random level generation is a "quintessential" feature -- I mean, you wouldn't say
Worms is a roguelike. I realise that most people would see him as an authority for having created Rogue, but that's no reason for us to be stuck with terms which render any discussion into a nonsensical one. I mean, imagine if he said the "quintessential" feature was that it's played on a computer. Heh.
3. His definitions for "dungeon crawler" and "Metroidvania" that he used to bash me are imprecise ("have lots of")... and wrong? A Metroidvania is not a "platformer that has lots of saving and backtracking". And if you argue that "oh, well, he didn't want to take the time...", well screw you! His writing should be held to the same standards that he holds the games he reviews.
Yeah, that was pretty vague and he's certainly not beyond criticism. However, it was actually
meant to be vague to demonstrate how vagueness can render a statement useless:
How can I explain this elementary concept to you more plainly? Here's another example: a metroidvania is a platformer involving lots of saving and backtracking. But many FPSes (or adventure games, or dungeon crawlers, or flight simulators, or, or, or...) also involve lots of saving and backtracking, yet no one calls these games "metroidvania FPSes" or "FPSes with metroidvania elements".
So you see, he's not proposing it as a definition, he's just telling you that stating metroidvanias "involves lots of saving (eh?) and backtracking" and then using it as a way to sort out what are and aren't metroidvanias, leads you to obvious contradictions. You seem to agree with him, so let's move on.
Regarding the critique of Spelunky itself, I'm definitely taking it into consideration as I want to make the XBLA version better!
Excellent, the most important thing to take away from his review is the criticism of the control system. Spelunker's NES controls are not meant to be emulated.

Make the game challenging despite the controls, not because of them. I plan to write my own Spelunky review, for what it's worth. I just can't seem to find the time to sit down and write it.
I chose the low-resolution and Game Maker the same way I choose a pen and a sketchpad when I want to develop some new ideas - it's faster and more fun for me. Now that I have the opportunity to put the game on a console, I'm "upgrading" to higher resolution artwork and "real code". If anyone claims I promoted the original game as anything else, they're making it up!
That's fair enough but a game using Game Maker will always be inferior to one using a natively written executable. Rapid application development is not a bad idea at all, though. The problem is many people will begin to think the "sketch", as you call it, is the "finished product" and their taste will degrade likewise. Actually, if you look at a lot of other "indie games" (and even some doujin games, like ring^-27) you'll notice that there's a trend of using low resolutions and low colour palette. Heck, even Capcom is guilty of this. Have you ever pondered the reason behind this trend?
However, I really like doodles, sketches, short stories, short films, etc. Sometimes they're as fun to look at as full paintings, novels, feature-length films. Sometimes they lead to larger, more interesting works.
Sometimes though, they remove the possibility of larger, greater and more interesting work from ever being developed since the taste of the masses is degraded as a result and no one would (or even can) put the effort into such things that we once enjoyed.
Terada has defined himself as a "rakugaki" artist, more of a philosophy than a style of drawing, in which one draws a little everywhere, all the time, without thinking too much, on notebooks etc.
I like that philosophy, obviously.
Well, can he really say it's "without thinking too much"?

There's a lot of thinking involved, even in doodling. It seems to me like an excuse to hide behind unfinished or otherwise-unpublishable sketches!
It's a misrepresentation of indie game developers that they treat the "indie games" label as an excuse to avoid criticism and avoid learning.
Perhaps not you, but a lot of people do use it as an excuse. Much like the people who say "it's a $5 game! Stop complaining!" to defend failures they've purchase, or worse "it's a FREE GAME why don't you do better" as a way to shut down criticism completely. In fact it brings us back to the "quadrupling of the pixels" issue -- an issue, that I'll leave for another time.
It's exactly why I make small games - to invite feedback and criticism to improve myself (and also to entertain, yes). I have no problem with people comparing Spelunky to Super Mario Bros. 3, for example, although if you wanted to be honest about my intentions, you would mention that I made the game as a side project for my own pleasure.
This is a great attitude. That said, if one were to criticise someone else's work here in the same manner, I've got the feeling that you'd jump in to defend them like the rest of the other people. That is to say, don't you think you hold yourself up to a higher standard than you're willing to hold other "indie" developers to?