I don't work at UR, so I don't know their process or thinking. However, I suspect the problem comes from specialization. My engineering work started just as microprocessors came into use. Even with a digital focus, I learned a little about analog. Today's circuits are paradoxically more complex and simpler than ever before. It is nothing for a consumer circuit to use millions of transistors. A good PC can contain over a billion. Such complexity was unimaginable in the early 1970s.
Simultaneously, today's designs are done with building blocks. A part like the MAX1473 is a complex circuit that took man-years to develop and a sophisticated fabrication facility to build. But you could buy copies of it for only a dollar or so and download a complete application note from the Internet for free. That app note shows you exactly how to use the part. Very efficient.
The UR system is a good one in concept and, to a great extent, in implementation. Someone has really thought through the feature and price points for various remote control capabilities and they've made a variety of receivers for those remotes. Whomever chose the Maxim part for this application, in my opinion, made a good decision.
In today's subcontracting world, once a circuit design is created, it is often jobbed out to a vendor for packaging and production. Things like bulk capacitor decoupling are 'tweener issues--they fall somewhere between actual design of circuit function and simple packaging. Sometimes such 'tweener issues simply get forgotten. There's huge pressure for low cost products and speed to market. Manufacturers simply cannot know how well a product will be accepted and hence try to control spending. Speed to market is important because if you don't do it quickly, someone else will. The consumer market, even for products in the price range of this stuff, is a difficult one to prosper in.
Testing is usually done by a fixed test fixture. In the MFX-250 case, the device works just fine in clean environments with short wires. The original engineers probably rigged one up and tested over long wires in a fairly quiet environment and got good results. Heck, they might have even had the bulk capacitance on their proto boards. I simply don't know.
These things happen though and they are harder to guard against than you might think. It is not a reflection on the intelligence of the designer. I doubt it was a conscious cost-cutting decision either. The loss of functionality and perception of unreliability certainly outweighs the small savings. I founded a couple of companies in the gaming industry. I'm an engineer by trade. Reliability was our utmost concern. We cared tremendously. A typical sale price was several million dollars for a system. And we had this exact same issue crop up on one of our product elements. Slipups happen.
This problem isn't limited to UR though. I just returned a $1,000 HP scanner because its USB port implementation was so finicky, it could only work with a select few computers.
One thing's for sure: as products grow more complex, as consumer pressure drives prices downward and as we use more subcontractors to maintain efficient production, we'll see more problems like this.
Would I buy UP products again? Absolutely! I am pleased with the functionality and the innovation.
By the way, I'm on my third day of operation after adding the capacitors. All is working perfectly.
