No there is no need for fixing the model. It does not have a problem. It's the same model of this: https://www.shapeways.com/product/9M4MGFYT4/vintage-camera-c ufflinks?li=search-results-1&optionId=42775119 Printed at least 4 times 4 pairs successfully. Vintage camera cufflinks. It was sold recently. It's the exact model just without cufflink back. Best regards
Shapeways personnel cannot know that essentially the same item in another file printed successfully. So they have to check it as if it was completely new, and as the design guidelines on the materials page cannot cover every case, some personal experience comes into play. So two people, or perhaps even the same person in different moods, may arrive at different conclusions when a part is close to the limits. In your case I wonder what the current thickness is - if they requested that you thicken the walls to 1mm it would follow that they are thinner than that, or they measured them incorrectly. Note that 1mm is the absolute minimum according to the current design guidelines - the 1.016 looks like excess precision from converting the equivalent value in inches. Either they used to allow thinner walls in the past and then thought better of it, or perhaps even your cufflinks "should have been rejected if the guidelines were absolute rules", although they proved to be printable. (And yes this topic comes up from time to time - last I remember was something like "n-scale car model" vs "set of 5 different n-scale car models" where the single-part files had been printed successfully. The only way out would seem to be to allow the designer time to respond before cancelling an order, but I can see how that might turn into lengthy discussions about "exactly the same" and "the same but may have been exported with different settings" etc.)