thinking outside the box

We can observe the cosmos on the past light cone and naturally, we see finite amounts of objects and of fluctuations which sets statistical limits on inferences from cosmological data (aka cosmic variance). Are there possibilities of looking outside the past light cone and what type of information would that provide?

bonus question: Can you estimate the value of \(\log\int_0^\infty\mathrm{d}t\:t^x\exp(-t)\) for large \(x\)?