Group delay
How long a frequency is delayed passing through the system, in milliseconds.
Group delay is the negative derivative of phase with respect to angular frequency: how long a narrow band of frequencies takes to get through the system. Every high-pass filter delays the frequencies near its cutoff, and a loudspeaker enclosure is a high-pass filter.
A sealed box is second order and delays modestly. A vented box is fourth order and delays about twice as much near tuning.
What is actually audible
Less than people assume, and the research is reasonably consistent. Commonly cited thresholds for audibility of group delay at low frequencies are in the region of 20 to 30 ms below 100 Hz: the ear is markedly less sensitive to phase at low frequencies than at high ones. Baffle flags peaks above 25 ms as worth a look, which is a conservative reading of that range.
Two things are worth knowing before worrying about a number here.
First, rooms do worse. Room modes routinely add tens of milliseconds of ringing at specific frequencies, swamping the enclosure's contribution. A vented box's group delay in an anechoic simulation is not what you will hear in a room.
Second, group delay near tuning is a consequence of the alignment, not a separate flaw you can design out. If you want the extension a fourth-order alignment gives, you accept its delay. A sealed box has less of both.
What it is not
Group delay is not "slow bass" in the sense the phrase usually implies. A loudspeaker does not lag behind the music by its group delay in any way a listener perceives as tempo. What it can do, at large values, is smear the attack of a low-frequency transient, and that is the effect worth listening for, rather than reading off a plot.
Every quantity in the workspace opens its own explanation where you are working. Open the workspace.