The only real reason it might be "wrong" is that it might be excessively expensive in the hear-and-now to do compared to how easy it will be in the next few generations. Although SpaceX and other companies are bringing down costs so quickly regardless that such a point might be moot.

Nevertheless, consider. Right now it costs on the order of a dollar per gram to put anything in orbit. And considerably more to maintain something like a space station in orbit, or to launch something to another planet. This is independent of the cost of the thing itself, which tends to be even more expensive. However, imagine what the fruits of improving technology could bring. Imagine if we could make 100% automated factories where bulk materials go in one end and fully constructed launch vehicles come out the other end (with robotics, 3D printing, and who knows what else in the middle). Imagine if we had computers and simulations so good that you could make a digital model of a launch vehicle and you could simulate launch and reentry with enormous accuracy, including exploring possible failure modes. Imagine if you then apply a genetic algorithm to existing proven launch vehicle designs with the goal of reducing launch costs and improving reusability as much as possible, and with the massive computing power of the future at hand you can pop out incredibly advanced and incredibly cost efficient designs for getting to orbit, and then build tons of them with your robot factories.

In short, in the not too distant future we could easily end up in a situation where building orbital spaceships is on the scale of building automobiles or airplanes, and it's as easy to get into space as it is to travel on airplanes today, or possible easier. That would make today's inefficient and expensive methods look like a waste of effort in comparison.

However, that's generally true for just about anything we do today anyway, so I think it's probably worth keeping at it despite the difficulty.