Sugar levels are perhaps the most interesting matter prompted by the February edition of Wine. I had set aside quarter of an hour to read the mag, which gave me five minutes, when I'd finished it in ten, to re-look at this matter.

The first occasion is the rather strange business of Havana Hills Sauvignon Blanc having its 5-star rating retrospectively stripped because it exceeded the 5 grams per litre residual sugar level stipulated for the competition. Well, according to some, the strangeness is it having been awarded the rating in the first place, as it's not really an ambitious wine: retailer Roland Peens in Platter calls it a "crowd pleaser ... lighter and sweeter than previous". Clearly also a competition judge-pleaser!

I confess I was surprised to read, in the comments of Havana Hills owner Kobus du Plessis, that the sweetness of the wine was not natural, but achieved with the aid of concentrate. I'd innocently thought that that practice was confined to the lowest levels of wine production in South Africa, the cheap mass-produced stuff, but I'm told it's quite common. Kobus du Plessis does make the point that dryness is partly a matter of balance, and that European legislation does allow for wines up to 9g/l residual sugar being considered "dry" under certain circumstances - but he neglects to point out that the basic rule in Europe is that a wine is dry up to 4g/l - not 5g/l as in South Africa. In fact, it was not many years ago that the rule was changed here from 4 to 5 - I'd assumed that this was to allow more latitude to producers whose fermentations had "stuck" at off-dryness because of high alcohol levels. But maybe I was wrong about that too.

Anyway, the point is the rules for the Sauvignon Blanc Top Ten competition required that the wines be dry, according to local definition, and the Havana Hills was incorrectly entered and correctly disqualified subsequently. Of course, it is quite intriguing to know how they found out that the wine's RS was in fact 5.6g/l rather than the 5 it was credited with in the magazine at the time, and which was presumably the deceptive figure given on the entry form.  Who told Wine mag, a month or two after the event, about the correct figure? The magazine just says the truth "subsequently emerged".... My guess is that it was a jealous competitor, who knew about the stipulation and had suspicions about this wine and then had it tested.

It's a good thing for the Chenin Blanc Challenge, however, that dryness wasn't a rule for entry, other wise three of the four wines scoring 4.5 or 5 stars would have been disqualified. The average residual sugar of those wines was 6.9g/l! The average alcohol was 14.3%, but that figure was skewed by the winner, Kleine Zalze Vineyard Selection 2008, having 15.18%. If you look at all the eleven wines scoring four stars or more, only one - one! - would really count as dry: the Spier Private Collection 2008, with an RS of 1.2g/l. That's what most people would call dry. The next-lowest figure was 4.3g/l.

At the awards lunch last week, we tasted the winning wine, as well as the "Best unwooded", which was the Simonsig 2007 (5.44g/l RS) and the "Best value", which was the Kleine Zalze Cellar Section (NB not Vineyard selection; this had a mere 3.4g/l RS). I'm pretty sure that at least many of the people drinking these wines with their lunch would, like me, have found the driest wine the most plausible accompaniment to food - especially on a warm afternoon. The overall winning wine was simply too rich, too powerful, and too sweet. Altogether too sumptuous, to use one of Michael Fridjhon's favourite words of wine-praise.

The point gets made over and over again that South African Chenin is great value - and so indeed it is. But people that like dry, rather acidic Sauvignon Blanc - as so many seem to at all levels - for their summer plonk, are not going to easily convert to this stuff. And I can't find it in my heart to recommend that they do. The show-winners are mostly show-winning wines, designed to impress with sugar, power and - usually - oak. And too many chenins from serious, ambitious producers are being aimed at the show-winners' table. They're not made for drinking - not by me, anyway.

I don't believe producers who claim that they can't get flavour from the grapes unless they pick ultra-ripe (with the result being high sugar and alcohol - I'm sure there's none of Kobus du Plessis's sweet concentrate involved here!). If it's true, they need to do a bit of serious work in their vineyards. Spier manages, as do many others.

Or is it all really about "show styles"? Is the final judgement on the Chenin Challenge going to be that it helped improve the quality of Cape chenin - to the point that it pushed the stuff to being undrinkably splendid?

• Full results of the Challenge are on the Wine website