Well the companies like Sierra Nevada that actually do add water to their wort after fermentation spend a lot of money on equipment to properly de-oxygenate their water before adding it into the finished beer…
NA beer is sweet because it is targeted at Standard American Lager drinkers, which is a consumer base that is generally not into hoppy beers. I think that the residual sweetness could be balanced via the addition of bitterness. While maltose is sweet, it is not glucose or sucrose sweet.
Relative Sweetness
Maltose Disaccharide 0.33 – 0.45
Glucose Monosaccharide 0.74 – 0.8
Sucrose Disaccharide 1.00 (reference)
Fructose Monosaccharide 1.17 – 1.75
I believe that the approach I outlined above would work better with a grainy tasting malt like Pilsner malt than it would with American 2-row. If we are simulating a green bottle beer, then apparent extract could be reduced. Green bottle beers are fairly well attenuated. Let’s say that the OE is 11.5P (O.G. 1.046) and the AE is 2P (F.G. 1.008)
RE = 0.1808 x 11.5 + 0.8192 x 2 = 3.72P
As calculated above, RE is going to be approximately 92% of the OE.
OE = 3.72 / .92 = 4.04P = ~1.016 S.G.
A 4P wort contains 40 grams of extract per liter, which means that one liter of low-alcohol beer produced from a 158F mash would contain 40 x .41 = 16.4 grams of maltose. Since maltose is 1/3rd as sweet as sucrose, 16.4 grams of maltose is roughly equal to 16.4 x .33 = 5.4 grams of sucrose in sweetness, which is equivalent in sweetness to about 1.4 teaspoons of sucrose dissolved into 1L of water.
Maltotriose is technically the shortest saccharide polymer that can be classified as a maltodextrin. Have you ever tasted maltodextrin powder? It is not very sweet.
I’ve had plenty of German NA beers that were still very sweet… I don’t think it’s because they are aimed there I think it’s due to process.
It is increasingly common for craft beers to be pasteurized, especially barrel aged ones after some high profile infections. It doesn’t seem to impact those beers negatively. I do not know whether that is because they are robustly flavored beers or if modern flash pasteurizing equipment works very well.
As for S. ludwigii, Fix also wrote that beers fermented with it taste like hopped iced tea (lacking characteristic secondary fermentation products as well as alcohol) and it was quickly abandoned by the NA beer industry.
Since nobody has mentioned this yet, there is an episode of The Brewing Network with Charlie Bamforth as the guest and there was a question about low alcohol beer. He gave the reality check that what industrial brewers produce is the result of millions of dollars of R&D performed by talented people and you are unlikely to do better in your home.
His suggestion was to mash very high (like 165) for a low alcohol beer and to not try to make non-alcohol beer.
Mashing in the 160s horrifies most people because they think the beer will be sweet but things that are not fermentable are either not sweet or much less sweet than sugar. Lagunitus IPA is mashed at 160 I believe and is not a particularly sweet beer.
S. ludwigii is still very much in use in the production of NA beer. It’s one of Hefebank Weihenstephan’s best selling cultures. This bank deals primarily with professional brewers.
After reading through this thread, my question is why, not how… 
+1. A good Mild is one thing but almost no alcohol - diminishing returns (to me).
Honestly if I could make a ‘beer’ with no alcohol that still had the taste of alcohol and all the other yeast byproducts… I would never make a beer with alcohol again. Obviously that’s not possible… but it would be awesome.
Well I for one don’t think “I would never make a beer with alcohol again”, but I do agree that a full flavored NA beer would be awesome to have available for those who can’t tolerate alcohol…