A couple of small disclaimers:
1) I
do think grouping is a good idea, but "it's complicated"
2) I am by
no means privy to the internal priority list for enhancements.. I only work to extend the list (to infinity), and I do not directly participate in priority setting.
3) I wish to offend no one, just adding my own voice.
From my own perspective, I don't see how it would be possible to add Size Groups quickly, nor economically.
For one, the current database does not seem to be setup in such a way as to support linked models. Every model stands (or falls) based upon it's own merits. Second, Shapeways would need some way to ensure that various sizes are actually "identical shape". Third, the FirstToTry feature, while imperfect, can produce positive results (setting customer expectation) quickly, possibly increase sales (or reduce the number of returns) and thereby funding the Grouping option we all desire.
Let's say I make a 2mm thick ring, and present it in various sizes from 10 to 15mm in diameter. That is six different shapes... but they are not
Identical. If the geometries were "the same", then the ring would be 2mm thick at 10mm, and 3mm thick at 15mm in diameter. Or, depending upon how my design process works, the ring might have started out at 2mm x 15mm and shrunk to only 1.33mm thick at 10mm. If I chose to print it in a material that requires 2mm walls, then the small ring at 1.33mm would fail and should not have been declared as printable just because the large ring worked.
In effect, I would be asking Shapeways to just "trust me" that the versions would all be printable, and their current systems require that printability is
demonstrated first. Since Shapeways is not aware of what my design process is, they must inspect each item individually. This inspection (perhaps) could be done as a paid-for service, but I don't think it's reasonable to ask them to just assume that two models I arbitrarily link together are both automatically printable.
===
Now, on the other hand, something I think Shapeways
could do (with less work) is to give the buyer the option to rescale a model upwards at the point of purchase (down-scaling won't work). If 'all' rings were designed at say size 5, then they could present the buyer with a dropdown of sizes 6,7,8,9,10 and use standard math ratios to increase the size. If they could do the resizing on their end, then they could know that the shape is identical, and larger items should "guarantee" to print if the smaller one did. Unfortunately, such a software fix would involve the walls on hollow items getting unnecessarily thicker as the item is upscaled. The buyer would incur extra expense over having each size designed individually to optimize the print cost. It's not a perfect solution, but it has it's merits.