That's a really good question.

Although I've never played with VOIP, I have been watching from the sidelines.

I have noticed that a typical cell-phone conversation has noticeable latency. It's bad enough so that, when I'm calling from my cell phone, I try to call a land-line if possible. I don't want to incur the full latency of a cell-to-cell call.

I have often wondered what the source of the latency is. Does one handset have worse latency than another because of variable processing power? Or is the latency dominated by delays in processing at the tower? Or somewhere else? Is the latency asymmetric? For example, is it mostly on the encoding side in the handset and not so much on the decoding side? It would be cool if an expert would drop by this thread to enlighten me. :^)

Anyway, if latency were a real issue for VOIP (and I have no experience to say one way or the other), then, even if the audio fidelity of what comes through were substantially greater than that of a land line, I'd still prefer a land line for a real-time conversation.

We cannot ever rightly say that general relativity is true, for it is a
scientific theory. A scientific theory is not something that can be
proved true, though it can be proved false.

In the hundred years or so since Einstein's introduction of general
relativity, no observation has produced data that would rule out general
relativity from its status as a candidate for the true description of
gravity. So in a loose sense, it still "holds true". But such wording
can be subtly confusing and, in my opinion, should be discouraged.

No experiment---not even any experiment related to quantum
mechanics---no experiment has exposed flaws in general relativity. It
is well known that general relativity and quantum mechanics are
incompatible. They cannot both be true. Still, no experiment yet
devised has been able to rule out either of those great theoretical
foundations of modern physics.

A theory of everything would need to be inconsistent, in some sense,
with either GR or with QM or with both.

That's a good point, and it should be elaborated as the proper response
to cyberchondriac.

cyberchondriac identifies the grid-bent-by-balls as "the current popular
gravity model". It is in fact a popular model, which I remember from
watching PBS even as far back as the 1970s. The good thing about this
model is that it allows one to visualize how a mass both distorts space
and moves in response to the distortion caused by another object. But
its goodness as a model of gravity ends there, in part due to
cyberchondriac's astute observation that it makes use of gravity to
explain gravity. Still, the model is not bad because it uses one aspect
of gravity (that it is nearly uniform near the surface of the Earth) to
explain a *different* aspect of gravity (that distortions caused by
multiple objects can interfere with each other and lead to motion).

Not quite. In reality, the best model that we have is general
relativity, according to which both space and time are being distorted.
But this is not to say that space and time are being distorted in
reality, because we will never know for sure what's going on in reality.
That is, a scientific theory (like general relativity) can never be
proved true, though it can be proved false. Who knows? General
relativity might be ruled out by some future experiments and replaced
with a fundamentally different view of gravity.