If you could restart your investment which Printer would you buy? (Page 1 of 2)

Actually, if I knew everything I've since learned by hacking my solidoodle 2, I'd probably build my own delta style printer (but I doubt I could have done it successfully without having the sd2 for a while - heck, I might not be successful if I tried now, but at least I'd be a lot more confident :-).

Have a Davinci 1.0 and just bought a Davinci 2.0 for more printing fun.  Bang for buck they kick butt.

Although I will not send another dime to Solidoodle, if I were to go back in time I think I would still buy SD for my first.  Yes, they are slow, their QC is MIA, and the printer is flimsy, but it was cheap and I learned a lot getting it to work (still am).  I see SD as a "kit that comes with the parts in their assembled position."  If you need plug and play - you'll have to spend some more bucks.

Re:Prusa i3: I don't care for the fact that the bed moves on the Y axis.  I think it is better to keep the model as still as possible - especially for higher speeds.  Also, although this is easily enough added for any model I suppose, I like having the enclosure.

For prints that aren't very tall, I agree with this.  However, as the print gets taller, I'm with TickTock. 

My take on it is that it won't be so much of an issue with quality, as it will affect how well a print sticks to the bed for certain objects.  Imagine a tall, "top heavy" object, for example.  I'm thinking that by printing fast, along with abrupt direction changes, would put more strain on bed adhesion, which can already be an issue with a "stationary" bed.

If the movement of the bed is enough to break the object free chances are the adhesion was poor to begin with. I've seen all kinds of models printed on robo3d/prusa style printers and what you're all discussing really is a non issue with a level platform and appropriate nozzle-to-bed distance. The bot will skip long before your object is knocked off the bed, if your adhesion is good enough.

Will you print PLA on blue painters tape at high speed with no heated platform on a machine like this? Probably not. But ABS, hairspray and a heated platform bond so strongly that it simply isn't a problem. People who have lost glass over a print that stuck too well know exactly what I mean.

What parts move to complete the build doesn't matter. What matters is mechanical rigidity and good adhesion. An advantage of machines that move the bed instead of the entire X/Y gantry is they can achieve mechanical precision at a far lower cost than the traditional Core XY/Ultimkaker style machines because the Y axis is just as simple as the X axis.

That is a tough question. I decided to jump in with a used SD$ . I have been looking for a deal on a flash forge. And I have been eyeing some of the delta kits.
There are a fair amount of choices in the price range you mention.
Prusa I3 kit
Kossel Delta kit
the solidoodle press or solidoodle SD4
Asterid plastic scribble 
Possibly a used wood box flashforge
prinbotot simple metal.

Tin

Same here.  I've thrown away two glass plates so far due to too many chips of glass missing from both sides (I use one until is becomes unuseable, flip it over and continue using until that side has too many chips, then toss it).  I've found I actually have better luck removing the part *while the bed is hot* - contrary to popular knowledge that letting it cool and pop off on its own is best.  Alcohol is helpful removing the glass chip from the part - doesn't seem to affect ABS but does dissolve hairspray.

Back on topic.  I do not know from experience but at high print speeds, I cannot imagine there *not* being vibrations at the top of a tall print impacting quality for a moving bed topology.  Do you just slow down on perimeters so the total print time is not impacted much or does the vibration I am imagining really not exist at speed (>60mm/s)?  For simple, less intricate, things with smooth features like vases I don't imagine it would be a problem either.

The simplicity of a moving bed is a big benefit.  I spent a lot of time trying to get rid of Y backlash on the SD and never completely succeeded.  My Ordbot has a sliding bed, and there is no backlash, since it is direct drive, just like the X.  I usually run outer perimeters slower than the rest of the print since that is the only place where speed/acceleration artifacts will be visible.

if you want to test this, just print a very thin and tall model on your glass bed and then shake the glass with your hand. Does it bend or pop off? if not, then there you go.

Hehe. It seems like it at first, and seems very counter-intuitive, but once you think about it it's actually not the case and is easier to set up (the moving bed, not bread ).

If the resin prices drop one day and build areas increase SLA will become more popular, until then I don't see it as a practical alternative for many users even if the cost of the printer itself is lowered.

I'd definitely think of getting one if I could make my own resin.