I've recently turned to the vitality starter advocated by Colin Kaminsky, who picked it up back in 2002 from a brewmaster at Coors England: Spin the yeast you'll be using for four to six hours in a starter on a stir plate prior to pitching, and dump the whole thing in. Do not oxygenate the wort.
This has worked very well for me, so I'll stick with it. I do not trust the SNS to supply enough oxygen to build sterol stores up to max, so if I use that, I feel I have to oxygenate the wort - and it's nice to be able to skip that.
Reusing yeast is great, and if you take the yeast from a healthy fermentation, you should have good glycogen stores. But those will be depleted fairly quickly in storage, so you should not wait long before using it. The breweries do not. You must supply plenty of oxygen, because the sterol levels will be low in yeast at the end of the fermenation, but with good stores of glycogen to fuel the synthesis of sterols, you'll have no worries. But that's also why you don't want to wait long before using the harvested yeast.
So if you can't brew within a few days from harvesting, I think you should refresh the yeast. Give it some 1.020-ish wort, spin it for some time, and then let it sit with an airlock to let it replenish the glycogen stores. It will take a couple of days. Then you might either use it as it is, but with oxygenation, or you can use a Kaminsky vitalizer (without oxygenation). The latter will give you a shorter lag time.
I feel that combining a refresher and the Kaminsky vitalizer optimizes your chances of a good fermentation. But that's just based on my own recent experiences with a series of pilsner batches.
Reusing means you'll not be using only the very best yeast - at least not when bottom cropping after bottling or kegging. I'm not sure how much of a problem that actually is, but I'm fairly certain it limits the number of times you can reuse the yeast. A solution might be to take some of the fermenting beer at hight kräusen - say 24-72 hours after pitching - and build up new yeast from that. That's advocated by the Weihenstphan laboratory, but might seem a bit too much trouble. I does to me, so I haven't tried it yet
.