Of course I am not a commercial wine producer but I have successfully done all three, i.e. field blend, juice buckets and post fermentation. I cannot say that there is a predictable difference in the three methods assuming one make judicious choices in varieties. Years ago we always did a "field blend" of Zinfandel and Muscat, which usually worked out well. I have blended juice buckets of Zinfandel and Muscat with good results. Probably the best wine I feel I have ever made was a blend of a very thick and heavy Brunello and a very thin Barolo blended at 75:25, both of which were at least a year into bulk aging.
Most of my early mentors had their own "recipes", e.g., 6 boxes Zin, 2 boxes Muscat, 1 box Ruby Cab, etc. Grapes came in by train from CA to Utica (NY). Each year the wines turned out at least slightly different, which is no surprise given that grapes are a natural product that varies by year. Plus we had no idea which vineyard(s) the grapes came from. Despite the lack of information, the better winemakers made a consistently good wine, year to year.
Keep in mind that field blending does not lend itself to comparison. Once the grapes are crushed, the decision making is done. Unless there are multiple blends, there's no way to judge how well the chosen blend turns out in comparison to anything else. That is simultaneously the beauty and the drawback of field blending.
Our 2023 wines used 3 techniques: selective blending, field blending, and Frankenwining.
Selective Blending: We planned to bottle roughly CF 60 / CS 30 / Merlot 10 and CS 60 / CF 30 / Merlot 10 blends, based upon a very successful commercial blend. The CF and CS were barrel aged, the Merlot in glass. Those blends did not work as the above ratios did not blend harmoniously. Based upon taste testing we completely changed the ratios, and the final result was CF 100; CS 90 / Merlot 10; and Merlot 55 / CS 35 / CS+CF barrel residue 10.
Why did the original blends not do well? An obvious reason is different grapes (CA) from the commercial wine (NY). The
varietals were the same, but possibly different clones and certainly different growing conditions. The commercial winery has been around since the 80's, and their vineyards date from the 70's, and it's 3rd generation family, so they know their grapes. Yet their blends are selectively blended, post-aging.
Field Blending: We fermented 2 Sangiovese juice buckets with the pomace from 20 total lugs of the CF, CS, & Merlot. It sounded good on paper and we decided to go for it. And it worked well!
Frankenwing: We had 5 gallons of CS/Merlot (aged together for storage space reasons), 2 gallons Merlot, and 1 gallon of barrel residue (cleared in the fridge). Given there was only 9 gallons (we'd get 1 case each), we simply Frankensteined it, figuring we'd either get a decent wine or a case each of cooking wine. Bizarrely enough, it turned out very good. Sometimes things work out well!
I realized a long time ago that there's no "one size fits all" in winemaking.