Bug: compose `ConformalMap` metric with player Rindler factor

[WrathfulSpatula]

(See also issue #58.) Here's the reasoning behind this change:

Firstly, grant me that we can calculate the local numerical metric, at a specific point, as seen by a static and stationary observer at infinity, if we have the apparent gravitational acceleration at this point for these preconditions, due to assumption of Einstein equivalence principle. (I won't argue this very hard, right now, but the point is that "a gravitational field is locally indistinguishable from a uniform acceleration.)

By whatever means, say we simply have the metric, as seen by such an observer. Then, the player perspective is embedded in the curved space defined by this metric. Then, to get the metric from the player's perspective...

1. We start embedded in the metric, at the position of the RelativisticObject.
2. We reverse this embedding, by multiplying by the inverse metric.
3. We can now apply the player-perspective Rindler-like deformation of the metric.
4. We re-apply the original embedding, by multiplying again the metric.

[WrathfulSpatula]

The ultimate point of this, to reiterate, is that we want to know the metric, not from the perspective of a static, stationary observer at infinity, but from that of the player embedded in the same background curvature. From the tangent space of the RelativisticObject, this effective metric undoes the intrinsic space-time curvature from perspective of static, stationary observer at infinity, applies the player's perspective, and then undoes the first undoing. (There's a small chance I'll flip the conjugation again, shortly, but this is on the right track.)