This is known as the Wait Calculation (sometimes referred to as the Wait/Walk problem) but it only applies to scenarios where a better option B, if done later, can overtake an inferior option A; for example a future spacecraft might be able to beat a less advanced craft to its destination, rendering the first a waste of resources. In the case of carbon capture though, the results are additive: whatever carbon an inneficient gen 1 plant removes before gen 2 is developed will remain removed when you go to implement gen 2, and any number X gen 2 plants plus Y gen 1 plants as well will always outperform just X gen 2 plants. Now if you were putting money (or some comparable resource) into escrow to be used for future carbon capture plants and the total resources allocated to carbon capture were fixed, it might make sense to save it for those better gen 2 plants, but for the moment no such system is in place.
Whatever money we spend on a Gen 1 plant can’t also be spent on Gen 2 plant. Whatever CO2 emissions we create while making a Gen 1 plant can’t be used again to make a Gen 2 plant. Whatever land we dedicate... Whatever ongoing power is used less efficiently for decades to come...
I think there might be ample valid reasons to think 20 years from the starting line could be better for waiting rather than racing to get an inferior Gen 1 plant online in the name of doing something. This is partly a political/social problem, but it’s largely an engineering calculation. “At the end of the useful life of both plants, which leaves the planet and humans better off?” should be the guiding question.
Again, it's only relevant if there is a fixed pool that both are being drawn from. Unless you are planning on spending 100% of world GDP and 100% of world land area on carbon capture plants, you are not sacrificing any future Gen 2 plants to build a Gen 1 plant. Even in the magical world where we know for certain to what degree Gen 2 can outperform Gen 1, and there was no cost to leaving carbon in the atmosphere for 20 years, Gen 1 + Gen 2 will still always outperform Gen 2 alone. In the real world, racing to do something now with technology that works is obviously better than letting the problem fester in the blind hope that it will become easier to solve later. Engineering is not about maximizing efficiency, it is about successfully solving problems. The guiding question is always "what get's the job done?" It doesn't matter how superior Gen 2 plants may be, if you can't build them they aren't an option.
