Hell no. We are not talking about a 100W CPU/GPU, we are talking about a touch controller IC that uses almost no power. Thermal cycling due to regular use is not an issue. I am not saying that the solder connection is not to blame, just that the cause of the problem is not thermal cycling. If one is having repeated failures then they are obviously carrying the phone in such a way that it bends. The back pocket is the worst place to carry a phone, but the front pocket can also be bad. Some people do not even realize they are doing it. But one thing is certain, if you have 9 successive failures, it's you. Better odds of winning a lottery then having 9 successive failures -- or at least it is close.
I noticed that the iPhone 7 is not any thinner then the 6+. A tiny bit thicker even. This bodes well for the durability of the 7 so it is possible Apple learned from their mistake. Not that the 6+ is defective, but it could definitely be stronger.
ARC results in much lower latency when cleaning up memory. This results in better user interfaces and helps explain why it works well with iOS. The JVM employs garbage collection that offers better performance at the cost of higher latency. If you load balance multiple instances (as suggested by another poster) you get overall improved performance. The JVM is better suited for server applications where its garbage collection implementation is safer and requires less overall energy to perform a given job.
Your problems with Android will have more to do with API design and your approach to programming said APIs then they will the type of garbage collection used by the platform. I am not suggesting that Android is great, but it has been shown to work well if approached correctly.
Tax cuts benefit those who pay more taxes - the rich. A family that exists just above the poverty line is not going to get anything back when they have their washing machine repaired. But a wealthy individual would get a significant amount back - hardly fair.
A better solution would be to increase the "environmental" tax and apply the additional tax revenue to pay for recycling and other environmental programs. Now everyone benefits equally. And by raising the price of appliances, the government would be promoting more repair and reuse of appliances.
The big change in DCDC design is in the different modes of operation that a DCDC controller can support. It used to be simple pulse width modulation but now we have pulse width modulation and, to use a term adopted by Linear, "Burst" mode DCDC converters. The purpose of the "burst" mode is to achieve low power level efficiency by on/off modulating the DCDC converter. The resulting on/off modulation can be within the acoustic range even if the actual DCDC converter is switching in the MHz range. So Linear, TI, Analog - they all now support their own version of a "burst" mode.
In the past 5 years, far more parts from various manufacturers are available for designing systems that goes to sleep but require always-on power rails. You used to have to pair a DCDC and LDO together to achieve the best of both worlds. And companies like Murata have capacitors specifically designed to assist in alleviating the whine. Check out their product line for a more detained description. I have designed and built power supplies that have had a noticeable whine - typically under low load. So I can confirm - it is the caps.
As of next Tuesday, C will be flushed in favor of COBOL. Please update your programs.