My understanding is that when you turn at the end of the timing period following a parallel entry, the idea is to intercept the inbound track to the fix - not track directly to the fix.
This gives you time to establish yourself on the inbound track of the holding pattern before reaching the fix and thus avoids the awkward manoeuvering needed to regain track after the fix - where depending on the final angle that you attacked the fix as you approached it after the parallel entry, you could find yourself badly placed on leaving the fix on final.
For example if you tracked direct to the fix rather than intercept track to the fix, you could finish up 25-35 degrees off track after passing the fix and having to S turn back to get on track again.