Anyway, I'll give the megastructure conjecture much more weight the second someone in the research community suggests it, preferably with some sort of a testable hypothesis. As far as I know it has been purely layperson speculation and as such not much more plausible than the ramblings of UFO believers or free energy cranks.
Well... almost.
It was originally suggested (as a highly unlikely but testable hypothesis) by astronomer Jason Wright. Then the press completely distorted his original point beyond recognition.
Methods for using Kepler data to identify artificial megastructures is the subject of research by several astronomers, including Luc Arnold (French National Centre for Scientific Research), Jason Wright (Penn State), and Geoff Marcy (UC Berkeley).
Here's a representative paper: http://arxiv.org/abs/astro-ph/0503580
Luc Arnold wrote a paper arguing that swarms of artificial structures would have certain unusual signature variations in their light curves. Jason Wright later pointed out that KIC 12557548 had similar predicted variations.
"Now, I don’t know what this is. Maybe it really is an evaporating planet (the best guess, I’d say). ... I’d bet my house on it not being aliens. But given that he went way out on a limb and predicted almost exactly this sort of thing, don’t you think Luc Arnold at least deserved a citation?"
This line, "I'd bet my house on it not being aliens," somehow got picked up by the media as "Astronomers have found super-advanced aliens," because that's how the press works.
So is it all lay speculation? Well, if you just mean credible astronomers aren't seriously arguing for this as a likely possibility, then yes. But if you mean there's no testable way to analyze Kepler data to distinguish between artificial and natural satellites... well... work in progress.
So given that aliens might exist, and this might be the most likely way for us to detect them, it's certainly possible that it really is aliens this time.
Of course we don't know for certain. Not by a long shot. It is absolutely speculation, but it's reasonable speculation. We probably need a lot more data before we can say anything more specific about this. Maybe we do find a natural explanation after all. But it's not reasonable to ignore this possibility because some people don't want to discuss it.
And the person who proposed this is a real astronomer. Of course he didn't say it's definitively aliens. He said it's a real possibility. That doesn't mean a lot (because it's technically true of any star or exoplanet), but it's the most likely case we've got. And that's worth some attention.
Neither of those things should increase the probability at all. Just because a researcher believes it doesn't make it more likely to be true. Neither does the fact that it's testable.
Conversely, you shouldn't lower the probability because those things aren't true.
A 1% probability of aliens seems pretty significant to me. Obviously it's more likely that it's not aliens, and no one disagrees with that. But the fact that it's even a possibility is very interesting and makes it worthy of investigation.
Anyway there isn't anything inherently unlikely about aliens. You make it like it's incredibly unlikely that aliens exist, so any hypothesis that includes them must be incredibly unlikely. A lot of people have higher prior probabilities on the existence of aliens.