Okay. It probably does, and I’ll personally consider using aspect markers when the meaning calls for them a best, most correct, practice, but I’m still not totally convinced that it’s definitely, 100% true.

Not 100% true. Just "usually" true. True enough that you should accept it as true, and take note when Okrand himself violates it.

To put it another way: if you can leave off a type 7 suffix whenever you want, what is the point of the text that says it's "usually" needed? Certainly it's not giving you a grammatical quota. ("You left off five out of your last ten aspect suffixes. Prepare to die!")

Just the mere fact that a verb taking a sentence as an object can’t have a type 7 suffix (but probably can have a perfective or imperfective meaning) makes me personally think that the suffix *may* not be totally needed to communicate that meaning. But it is also entirely possible that taking a sentence as object is the only case in which these suffixes may be omitted despite the meaning calling for them, or one of a very small number of other cases.

Another case is the rule of "never" putting a type 7 suffix on a verb with -jaj. (Except he has. Does that mean the rule can be ignored whenever you want? No. The rule is the rule, and there is some reason it didn't get applied to that particular sentence. Error? Figure of speech? Allowed at night on a Tuesday? We don't know.)

Anyway, I do think the language in TKD 4.2.7 certainly suggests that using the aspect markers is a good thing to do, but I’ve seen a lot of (non-canon) usage that seems consistent with a looser, more “optional” view of the suffixes than the one you promote. I haven’t yet studied the canon sufficiently to see if the same holds true in canon.

Canon nearly completely supports my interpretation. Most of what Okrand has translated has no need for perfective, being proverbs, descriptions of objects, and storytelling. When he does need perfective, it's usually for the speech of people talking about completed events or the background of people or objects, and he uses it fairly consistently. You'd be hard-pressed to find a verb that really ought to be perfective that isn't. And I'll bet most of those you do come from his earlier stuff, when he made more mistakes.

There are violations of the rule: stuffed tobaj leg is to'baj 'uSHom lughoDlu'bogh, without a needed -pu', but Okrand has used a lot more -lu'pu'bogh and -lu'ta'bogh phrases, and the one other time I know of when he used a -lu'bogh verb, in paq'batlh, it was describing something that wasn't complete, so it didn't need a perfective suffix.

In particular, the Duolingo course seems to use verbs with no aspect markers with English translations in the simple past quite regularly, in sentences where it seems like the meaning would indicate a completed action. It does seem that at least a few people feel that the aspect markers can be left off, so I’d be interested in hearing some arguments in favor of such a view as well, if anybody has them.

For decades members of the KLI explained Klingon perfective as meaning "happens before the time context." I know, I was one of them. Everything we translated or wrote used (or didn't use) perfective that way. The trouble is, it's wrong. "Before the time context" is tense, not aspect. When you describe when something happened, instead of the way it happened, that's tense. And Klingon doesn't have morphological tense.

The old explanation completely failed to explain canonical phrases like loSmaH bej jIboghpu' I was born forty years ago. We had to twist ourselves into pretzels, saying, well, I wasn't born exactly forty years ago, so as of forty years ago I had already been born. Which is a useless and pointless thing to say to give your age, since as of thirty-nine years ago I had also already been born, and so on. This particular phrase was the first time I doubted the traditional wisdom of the meaning of Klingon perfective, and I've been studying it ever since.

As for Duolingo, it carries no especial "correctness." It was created by people on this list, indeed probably reading this message.

The "usually" is just part of Okrand's usual bit about the dictionary being only a basic sketch of the language.

Which in turn is probably because he didn’t want to pin things down in too fine detail, to leave some flexibility for future work on the language.

I don't think he expected there to be future work on the language when he wrote that. He expected the book to sit on the shelves of some Trekkies, and they might learn a few phrases. He didn't pin it down because it would have been boring to put in the book, and verisimilitude demanded he make it seem like the language wasn't well-understood anyway. So all the given rules are wibbly, but they're generally true.

-- 
SuStel
http://trimboli.name