Thanks for the feedback. It looks like you are aware that this is not the most appropriate avenue for this discussion, so you might want to think about what prompted you to post here before trying to contact us some more conventional way. You do bring up two subjects that I think are worth discussing publicly.
First and most directly addressing your comment, you got quoted a little over $500 for over 2500 identical little parts. A laser cutter is not necessarily going to be the most cost-effective way to do that kind of job, and if you have a better way of doing it, great: do it that better way. It's a similar situation with the competing quote: I don't think we promise being the lowest-cost option, but you might not be getting an apples-to-apples comparison since there are other factors like turn time, quality, support, etc. I'm not saying the cheaper options are necessarily going to turn out worse: I might have done your job for half or less back when I was running a laser out of my house. Still, a lot of your ire seems to be stemming from your disbelief about it taking four hours, and that does not sound that off to me.
Second and more troubling is your resorting to complaining about greed, which seems all too popular lately. There's some rate at which things are worth it for us to do, and if that rate does not make sense for your application, don't go for it. There are probably things that you would not do for $100 but that you would do for $1,000; does that make you greedy (in some bad way)? You say, "Might as well get them all cut, right?"; isn't that the same greediness? It's fine for you to come back to us with your other quotes to make sure there wasn't some miscommunication about your requirements or to try to negotiate a better price, but it's not appropriate or helpful to stoop to that kind of name-calling. Also, you should want to work and interact with competent people, for whom "greedy" probably isn't a particularly meaningful word.
The Maestros should give very good, smooth outputs, so the jerkiness is probably because of the servo you're using or the load you have on it. You're probably effectively getting a situation like the one in the last two diagrams, where the commanded position is gradually changing but the servo responds in jerks either because it isn't very good or because you have so much load on it that it can't move until the error is quite big. Can you put the same servo setup on an RC receiver and slowly move the controls on your transmitter?
I want my office near Pololu Valley! We have customers on the Big Island, so it might be possible to see some of our products in action there (I remember hearing some talk of a robot club in Waimea a while ago, but I don't know what came of that). I don't think we have any distributors in Hawaii.
The main motivation is cheap, relatively easy control: you can use just a single I/O line on a microcontroller, and you don't have to deal with any power electronics to drive the motor. If you want a motor controller that is more than just an H-bridge that you might have to do a lot of work with, it's difficult to get that and the motor and gearbox for the ~$10 of a cheap servo.
The servo is not made for continuous rotation, and it rotating continuously is a sign that you are asking it to go beyond its limit, which could cause it to be destroyed. The standard 1-2 ms range should get you close to the full turn, and you can gradually expand the range while keeping it centered at 1.5 ms to get to exactly one rotation of range.
By the way, you should direct questions like these to our forums or to more direct tech support.
I don't know how you expect me to be able to judge that better than you. However, from the little information I can get from your question, it sounds like you aren't getting the point. This article is about a specific approach to getting many channels of servo control. If you just need one or a few servos, you don't need anything this complicated, and if you need to do many servos, you're probably better off getting a dedicated servo controller.
It looks like you might have already asked about this on the forum some, and I don't really have a good answer. I would look around for different servos (you can get metal gears with analog servos, by the way), but if you like the ones you have, you'll have to go for the external power switch route you described. Some of the fancier servos are programmable or configurable, so you might check into whether the no-signal behavior of your unit can be changed.
My comment above already covers the easy part: you get a simple tradeoff of speed for torque, so, if you make the motor give you 20% of its stall torque, it will give it to you at 20% less RPM than the no-load RPM. However, calculating how extra weight on the robot translates to torque on the motor is not something you can easily calculate since it depends on things like friction and how the weight is distributed.