Ian Sutton wrote:The thumbnail generalisation I've seen argued most often is Magnums having a 50% longer drinking window than normal bottles. Extrapolating this down to halves, you might guess at ordinary bottles having 50% longer window than halves (i.e halves have 100/150 = 2/3 the drinking window of normal bottles).
FWIW I've personally thought this generalisation a little extreme in it's differences, but I have no great personal evidence to go on to contradict it.
I agree that seems a little extreme. I'd say 25-35% difference longer for magnums seems more accurate, extrapolating to 375s drinking window being maybe 3/4s or so of 750s. But like you, I'm working on limited anecdotal evidence.
As noted, older half-bottles of wine might have provenance issues, especially as they are not the size of choice of collectors. Older halves might well have been bought at a restaurant bankruptcy auction, for instance. Still, I occasionally take flyers on older halves, and have had surprisingly good results with '83 Beausejour Becot, '91 Ridge York Creek CS, '78 Lascombes, and other bottles where survival in 750s wasn't a slam-dunk.
I like halves. They're great for a romantic dinner for two with differing needs for main and appetizer, they're perfect for nights Betsy works (she'll usually have maybe a 2-3 oz pour), leaving me with 2 generous glasses. I wish selection was better. I also regret that sometimes the pricing is such that it seems stupid to buy a half ($14 375, $19 750, even if I dump a third I'm ok with bigger bottle).
Oh yeah, and I buy 90% of my dessert wines in halves.