NEW YORK - Earlier today, General Motors announced that Chairman and Chief Executive Rick Wagoner has taken control of the company's troubled North American operations. GM Vice Chairman Robert Lutz, who was rebuilding the company's car and truck lines, and Gary Cowger, who was president of GM North America, have been told to concentrate on global problems. This isn't the same as moving deck chairs around on the Titanic. GM's (nyse: GM - news - people ) latest move is more analogous to the ship's purser taking the wheel. Wagoner headed GM's North American auto operations before becoming chief executive. If there was any improvement during his command, I missed it. In addition, by his own admission, he isn't an expert on cars or trucks. He's a financial executive. GM traditionally has had someone with a strong financial background run the company, but there's usually been a "car guy" near the top. A few years ago, Wagoner hired Lutz, a seasoned and well-regarded auto executive, to be GM's car expert. I think that Lutz has done a fine job of starting to rebuild the product lineup. The new cars are much improved over the old ones, and more are coming. Given GM's problems, people are starting to say desperate things--such as suggesting that the company kill one or more car lines. Lutz inadvertently raised the subject, too, in a meeting in New York City during the Auto Show, and now everyone is on the bandwagon. Even a recent New York Times column on the advertising industry gives advice from--heaven help us--ad men on what GM should kill. First, let's clear up the misconception that there are eight hungry GM divisions, and that eight lines are just too many to feed with existing production. Let's start with Saab. With only 40,000 U.S. sales, Saab is too small to be an American problem. It's a European problem. Besides everything else, there aren't going to be genuine Saabs anymore. From what I understand, all of Saab's future models will be based on platforms from GM's Opel division in Germany, or GM's Subaru associate in Japan or on trucks from GM's U.S. factories. Maybe these new models will help increase Saab sales. Maybe they won't. Either way, feeding Saab isn't going to overwhelm GM's American operations. GM's Hummer operation is also too small, and not a serious problem. Its best-selling model, the H2, is now good for only about 30,000 sales a year. Yes, I question the wisdom of sticking a wimpy five-cylinder engine in the upcoming Hummer H3. I'd like to know who at General Motors (nyse: GM) figured that consumers who like big, bad Hummers would find this engine attractive. But this smaller Hummer should help improve sales. I also don't see the GMC division as a problem. It sells big SUVs and pickups, and that market is still viable. GMC is a great nameplate, and 1,000 dealers sell both GMC and Pontiac vehicles. Saturn is GM's most precarious car division. It has been starved for products but has stayed alive, thanks to its excellent dealers. A few new Saturn vehicles are on the way, so I think Saturn will be in better shape in a few years. Other people are offering advice to kill Buick or Pontiac--or both. Their theory is that GM could concentrate on its remaining product lines. If killing lines of cars or trucks or closing factories were all there is to the auto business, anyone could run a company. Who can't say: "Kill those cars, shut that factory?" But the art is in building attractive products, not killing nameplates. Pontiac sold 474,000 vehicles last year and is getting more models: The Pontiac Solstice two-seater comes in June and the small Torrent SUV this fall. Additional models of its G6 passenger car are also on the way. These additions are not enough, but they are a good start. We all know what Pontiac should be: an exciting driver's car and an affordable alternative to a BMW. Yes, Pontiac has to do a lot more work to reach that goal. Buick sold 310,000 vehicles last year. The new LaCrosse isn't breaking sales records, but is much improved over the models it replaces and will earn Buick some respect. The new Buick Lucerne--a big car with either V-6 or V-8 engines--will be here by the end of the year. And we all know what Buicks should be too: big, upscale cars and SUVs with lots of room, power and flashy design. Softer riding too. Unfortunately, GM just killed the plan for a line of rear-drive Buicks. That was bad news and a setback for the division. Buick and Pontiac still are short of product, but they account for 17% of GM's U.S. volume. Kill them and you lose 800,000 annual sales and scare off hundreds of thousands of other buyers, who figure that all of GM will be going out of business soon. So kill just one line instead of both? Some pundits think that Buick should be put out of its misery. But those people should go back and look at what happened when GM closed down Oldsmobile a few years ago. GM sales, market share and profits did not improve with the death of Oldsmobile. And as far as being able to put more effort into the surviving nameplates, just how much more effort did GM put into Buick when it killed Oldsmobile? Those with an appreciation of history should note that Buick predates the formation of General Motors and was GM's first nameplate when the company was incorporated in 1908. If GM can't save Buick, it can't save anything. The point is that killing divisions won't free up money or manpower to help the survivors. That's a myth. And the way GM works, the more vehicles produced off a common platform, the more it can spread out the overhead costs of tooling and bureaucracy. Many dealers already sell Pontiac, Buick and GMC all together. The dealer force is GM's best weapon today in the auto wars. As I've said for a long time: The key to success is to build better cars and trucks--not to kill nameplates.