Here's my little mini-site:
This is a holiday promotion aimed at my customers. They're overwhelmingly non-technical folks, and not too sophisticated at the Internet. (I am well-aware that most people here would not purchase my software. I'm totally OK with that: you are not the audience.)
I asked the Turkers three questions (and will tell you my rationales):
1) Is it IMMEDIATELY clear what the site offers when you open it? [Rationale: This site is a sledgehammer, not a scalpel. I want them to understand what it offers before they hit their back button. If thought is required, I failed at design.]
2) Try to print out Christmas bingo cards. Tell me how long it takes you from me giving you that instruction to something coming out from your printer. [Rationale: My target is that fourty-something mother with kids, with no special computer skills, can accomplish this in 90 seconds. There is actually a bit of an overlap between Turkers and my market.]
3) If you thought your friends would find Christmas bingo to be a fun activity, would you recommend this site for it to them? Why or why not? [Rationale: The Turking population has a significant overlap with my core audience. Why not ask a quick thumbs up/down?]
What it taught me:
1) The site achieves its singular purpose fairly well.
2) It is attractive enough to move several testers into mentioning it to their friends. (As in, not hypothetical.)
The feedback lets me know "OK, the site is basically OK and needs to be built-out more than rethought from scratch". Knowing that, and the peace of mind from knowing that, are worth $7 to me easily. I hope this service continues getting developed, because I can see a lot of uses for it in my business.
(Yeah, I could roll my own with the Turk API... but I value my time at $100 an hour. This means I never want to be playing the sit-at-computer-refreshing-Amazon-so-I-can-authorize-fourty-cent-payment game.)