I think there are several problems in conlanging, and I think they appear on several levels. I think the best way to explain my thoughts is to compare and contrast with earlier statements in this thread, so here I go:
Mostly because it nearly never rises past the status of a hobby. People conlang in a lazy way, and I'd say it's fair to say that over 90% of all significant conlangs end up nowhere. There IS a way to make conlanging into something more substantial than just a hobby. Why can't we make it art?
Is it really laziness that is the problem? I suspect the problems rather are due to lack of imagination and also lack of knowledge. These days, when there's a billion papers on topics in linguistics available for free online, the latter is perhaps the least excusable problem. I find increase in knowledge also tends to help imagination flourish.
The fact that very few people seem interested in reading about what you do shouldn't matter. It's like, if you don't get raving reviews for your 1 page phonology, you give up immediately. Conlanging is about making languages that WORK. It's as if all amateur mechanics only ever drew shitty drawings of shiny red cars but never built anything. I'm sad because that's the best part of conlanging, when you can see your baby develop into something you never would have imagined.
I think part of the problem here is that very many conlangers start out by religiously adhering to the order put forth in the LCK. If someone new asks how to go about constructing a language, that order also often is what is recommended. Alas, this means phonologies will be the bit every new language begins with, and so there will be a huge lot of them. Would you be as annoyed if the drafts that never get anywhere more often were, say, descriptions of how the case system and the verbal aspects of some language interact? Descriptions of how deixis is treated pragmatically? A system of deference? These are the kind of stubs I am going for at http://miniatureconlangs.blogspot.com
I'm just annoyed at the HUGE amount of shit being produced that's called conlang (usually half-assed phonologies). I wish conlangers put more effort into what they do, and most of all, I wish conlangers would fucking finishing what they start once in a while! Even if it takes years!
One problem here is that in nature, languages are huge things. Really crazily huge things even. A human could probably never create something as intricate as a natural language. On the other hand, paintings and sculptures never perfectly reproduce the thing they depict - they simplify and catch some or many of the visual essentials (or caricature them). In creating a conlang, we have to simplify - but we probably want to catch enough of the real thing to have it workable, of course, at least for the prototypical conlang.
However, I think a problem here is that many conlangers just don't realize how much there is to a real language- there's any bunch of idioms, almost every word will have connotations, there'll be variation in how the speakers use grammar and words, and with a big enough population this usually also will encode sociolinguistic data, there'll be poems and short stories that anyone wanting to properly understand the discourses in the language may need to know to be able to appreciate all of it. Languages are unfathomably large things to fully capture, let alone create.
Once this realization hits you, you may realize that some kind of not-quite-complete language is the only thing we can ever hope for.
On the other hand, I hope the idea that my blog stands for - minimalism, and downright avoiding phonologies - might be of some interest. I hope to develop some of the ideas deeper, and at some point maybe posting a bunch of ideas where I also provide some sample texts illustrating the idea, not just showcasing them using glossing.
I have more things to say, but I realized it's getting late so I will probably write more tomorrow or friday.