Upgrading tyres is one of the three effective ways to spend money on increasing your speed in any meaningful way.
(The other two being related to aerodynamics:
• buy a new jersey — or have your own tailored — so it fits snugly, doesn’t crease or flutter in the wind, and
• buy a narrower handlebar — considering that the pre-installed tend to be too wide. This is especially true for “gravel” bars. Of course, that is assuming a separate bar, and not an expesive carbon barstem — in which case the efficiency would be not
And no, making a bike lighter is far from efficient. Even in the rare occasion of a significant weight shedding for cheap :)
So. There is bicyclerollingresistance.com
— a helpful resource where you can find exactly how many watts a particular tyre will save compared to your current one (or competing for your choice).
However, the website’s built-in tyre comparison tool doesn’t take into account a couple of factors, either or both of which can lead to a considerable margin of error.
1) Tyres are compared at the same pressure (at same width) or equivalent (at different widths). However, tyres of even exactly the same width can differ in optimal pressure for the same weight by ~10 psi / 0.7 bars – depending on how stiff their carcass is. This, in turn, produces a discrepancy of about 10 watts per two wheels (according to their own test protocol: 85 kg at 29 km/h). Just the difference between the mid-range and top-end tyres.
2) Rolling resistance is measured at pressures that are selected by tyre width cohort. These cohorts are arranged in such a way that if two tyres are only 0.6 mm different in width, they may fall into different cohorts — in which case they will be compared at pressures differing by up to 6 psi / 0.4 bar. On the other hand, the opposite is possible — tyres with a width difference of as much as 3.9 mm could be placed in the same cohort and compared at the same pressure.
Combined in a least-favourable scenario, these two factors can produce an error comparable to the difference between the slowest anti-puncture tyres and the fastest racing ones. Not an ideal situation, especially considering that the data itself is very good, complete and comprehensive.
Plus, everyone always talks watts based on the mentioned protocol: 85 kg at 29 km/h. But if your weight is different from 70-something (after dividing bike, clothes, helmet, shoes, water) and your average speed is different from the suggested, it’s difficult to compare how fast the tyres will be in your circumstances.