I have only worked with lees on red wine, the result was good, but there are many factors involved and I haven't done any side by side studies to make strong conclusions. The main concern with lees stirring in red wine is potential color loss. I followed Clark's recommendation to remove the fine lees and reserve in a separate vessel, and allow the wine at least 4 months for the color to become bound. During this time, the vessel of lees need to be stirred regularly and have access to small amounts of oxygen to prevent them from becoming stinky; in my case I used a gallon screwcap jug with about 60ml of head space, this way every time you open to check for off odors the headspace is replenished with air then recapped and shaken to suspend the lees completely. After the 4 month period, as long as the lees are clean smelling, they can be added back to the wine and stirred. The process is a PITA and the winemaker has to determine if the result is worth the effort.
Regarding white wine, below is a response from Clark Smith on a comment and question about his Faux Chablis. I suppose it just highlights the complex issues with winemaking, from start to finish every single step affects the next, then it takes multiple years to determine if you achieved the desired results.
Dude, the smoke in that Faux Chablis
just sends me. You say it's yeast autolysis. I've stirred a lot of yeast, but I never got THAT! Is there a secret combination of elements? Or maybe I didn't stir often enough long enough?
Mark:
Thanks for the nice comments on the Faux. We're trying to show two things in this wine. One is that California Chardonnay doesn't need to be an oaky, toasty butterbomb -- blame the winemaker, not the terroir. This wine shows the distinctive lemon oil character because the alcohol isn't very high (12.9%), lowered from 14.8% original at dryness to a sweet spot. A high degree of ripeness is essential, just in Chablis, to get this character, but in CA wines it hides beneath the alcohol, and we have to adjust it. In Chablis, this isn't necessary. Instead they adjust the alcohol UP with beet sugar to make up for the dilution from rain.
The other is that white wine can achieve soulfulness through structure, same as a red. To build the structure, we first ferment on well cured, untoasted Alliers chips from Boise France. Then we need to complex the wood tannins with lees, the same way you complex Chenin Blanc's natural tannins in a sur lies Savennieres.
The smoky character is a product of prolonged slow transformation of the lees. It’s a little like marmite, and probably involves slowly evolving low temperature maillard reaction products. I’ve never seen them in batonage wines until at least three years later.
These wines are quite steely when they're young, and only open up after four or five years in the bottle. Worth the wait, though.