ForceSimulation does not update its alpha prop based on the internal d3 simulation

[regexident]

By not updating its alpha prop there is no way to re-heat a ForceSimulation by resetting its alpha back to 1, once it's already at 1, since Svelte's reactivity will swallow the re-assignment.

Apart from this as a user one would also expect alpha to reflect the current state of the simulation's alpha value.