Player::try_seek docs are wrong about only blocking up to ~5ms

[liquid-dragons]

This is stated here:

rodio/src/player.rs

Line 230 in ee3c21f

/// This blocks between 0 and ~5 milliseconds.

Which is referring to the periodic access here:

rodio/src/player.rs

Lines 137 to 138 in ee3c21f

	// If you change the duration update the docs for try_seek!
	.periodic_access(Duration::from_millis(5), move |src| {

However, compare this to the documentation for periodic_access().:

rodio/src/source/mod.rs

Lines 547 to 558 in ee3c21f

	/// Calls the `access` closure on `Self` the first time the source is iterated and every
	/// time `period` elapses.
	///
	/// Later changes in either `sample_rate()` or `channels_count()` won't be reflected in
	/// the rate of access.
	///
	/// The rate is based on playback speed, so both the following will call `access` when the
	/// same samples are reached:
	/// `periodic_access(Duration::from_secs(1), ...).speed(2.0)`
	/// `speed(2.0).periodic_access(Duration::from_secs(2), ...)`
	#[inline]
	fn periodic_access<F>(self, period: Duration, access: F) -> PeriodicAccess<Self, F>

Since the access rate is based on playback speed, if I play my music file at half speed, try_seek() I could be waiting for the next access up to ~10ms (or even more for lower playback rates).

---

This is related to, but technically distinct from these other two seek issues I've found: #638 #876.