Don't know the first thing about electronics, but I am interested to learn more about what you meant here. Highly-optimized code (usually) isn't too readable or extendable, are you saying Intel has the same problem with their processor's design?
Then these tradeoffs tend to multiply each other. If you have a design which is fast on branch-predicted and slow on branch-mispredicted, and you build a big branch predictor, it works well. But you're committed to powering all that.
I mean, compare http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.... (few sentences) with http://www.agner.org/optimize/microarchitecture.pdf (admittedly a lot more detail, but it's clearly a more complicated system too)
Large amounts of speculation makes sense for data center CPUs, but not for battery-powered devices.
Maybe it's more about peak performance vs. average power consumption under very low load as you usually have in a mobile phone that's only used occasionally. For example when starting an app and scrolling you need very high performance but the rest of the time the CPUs are mostly idling.