This really needs to be a sticky somewhere whenever folks bang the drum about Broad Vintage Generalizations. Here is a quote from a very wise man who taught me a lot about wine via the wine Internet when I got started way back when:
I think I don’t share your assumption about vintages being just better or worse (I bang my multivariate drum again). Vintages are different. If you have to assign points to them, you get a linear scale, but that doesn’t capture the different utility of, say, 2006 vs. 2007 Puzelat romorantin. You would serve them with different food. Maybe at a different time of year. But you don’t have to call one “better” in an absolute sense and think that you mean it. 1928 is a better vintage of Leoville Poyferre than 1929 today, because the '29 is OTH. But probably for the first 40 years you would rather have had the '29. '05 is a “better” vintage in a bunch of Burgundy than '04, unless you want to drink the wines anytime soon. Similarly in Bourgeuil. The reverse in Muscadet.
Variation in character may not imply variation in quality, at least in regions where I drink a lot of wine. Quality isn’t univariate. And so on…
The variation is interesting and appeals to the geek in me. Sometimes it is qualitative to the point that the vintage was like '80 in Vouvray, and it’s really tough to drink. Or it’s like '02 in Vouvray and everything is bliss. But sometimes it’s just interesting to have an '03 Haut Lieu non-greffe demi-sec and think, gee, I don’t love this over the '02, but what an interesting animal to have to compare. Maybe '02 is “better,” fwiw. But maybe it’s nice to have a few '03s in the cellar too for the light they shed on the '02s (and the '80s).