Interesting. If McClellan had McDowell from the beginning and they attempted that turning movement on Yorktown from Gloucester Point could you see the siege ending much earlier? How much impact would this have had in the weeks ahead.
It would have been difficult because of the transportation requirements - from memory McClellan was expecting a lot more transport than he actually got (he was expecting about 5-6 divisions of personnel lift, he got one division and change) and as such it was fully occupied moving his troops into Fort Monroe.
If he'd been able to make the landing at Gloucester Point, though, it would have turned the position - chiefly because it would now be possible for batteries to be set up on Gloucester Point and take Yorktown in enfilade (the redoubts are a mile to two miles from Gloucester Point) and threaten a second amphibious bound across behind Magruder. This would lead to the abandonment of the Yorktown line (OTL it was abandoned when McClellan was about to begin his cannonade, so a cannonade from Gloucester might not be necessary) and might save as much as 2-3 weeks.
The rainstorm of OTL might cost some of those days though.
More importantly, however, without the need for McClellan to hold his OTL position on the Chickahominy for weeks to allow McDowell to come down and reinforce him (as he'd already
have McDowell) then McClellan would be able to conduct the campaign he wanted - to use the James river as his base, and to push the Confederates back into Richmond and Petersburg before conducting a battle of posts (i.e. to advance his lines one step at a time under the cover of heavy guns, functionally forcing the Confederate army to either storm his lines (taking heavy casualties) or surrender the city). Petersburg would go first as it would open the James River (by defeating the batteries at City Point) and then McClellan and the gunboats can get practically into Richmond just by following the river.
This would be no Petersburg campaign of OTL, I hasten to add, because McClellan knew how to conduct regular approaches - how to suppress the enemy batteries with his own heavier guns, and then have an infantry attack go in under that cover. That's what he was planning to do at Yorktown, after all - suppress the enemy guns and then mount an assault.
If we assume he'd reach Petersburg by the end of May (which is extremely ungenerous to him, as OTL he was before Richmond by the end of May even with the delays and Richmond is further) and that the preparations for the storming would take as long as they did at Yorktown (i.e. 16th April to 5th May OTL, about three weeks) then you'd see Petersburg taken during June and the approach to Richmond up the James taking place over the end of the month - if it's not been evacuated, of course. If it's still occupied, then McClellan has a fairly hefty numerical advantage and a second attempt at regular approaches would do nicely.
Note that this is largely a set of pessimistic assumptions for McClellan, as I've assumed that the turning movement doesn't save
any time! If it does save 2-3 weeks then the Petersburg siege roughly takes the second half of May and the first part of June, and the Richmond one would conclude by July.