Like VR, touchscreens have not good enough latency. Because many components and their architecture contribute to latency (touch-CPU-RAM-app-OS-GPU-display), an

Like VR, touchscreens have not good enough latency. Because many components and their architecture contribute to latency (touch-CPU-RAM-app-OS-GPU-display), an integrated approach better improves it. NB: For Christensen, "not good enough" does not mean "not adequate", it means that even better would be valued. It means "need not satisfied". Consider a meal: too small, you're still hungry (you'd pay for more); too big, you leave some on the table (your need has been "overshot", you can't absorb the extra improvement, you won't pay for more). However, smartphones have overshot the need for pure computing power, it seems to me. BTW: I too find his disruption more compelling than modular/integrated. A theory is not very predictive if there's many ways to apply it, giving different results.