I've got a road bike and a gravel bike. A while ago, I did the same workout, 12 reps of a local hill, 2.5 hours or so, once on each bike, separated by a week.
The road bike is 8kg and 25mm tires, the gravel is 12kg and 650bx42 or 48 tires. (might have been either, as I switched tires around then). The difference in time was 6 seconds.
The gravel bike doesn't feel as fast. It doesn't reward spinning, and the gearing jumps are just a little big for me (1x11). The road bike beats me up with vibrations, but it feels better climbing and standing. It's got closer gear ratios, so it's just a bit better matching. Now, to be fair, the gravel bike is running far better tires, so there's less resistance there. (and, that's basically why I bought it) But overall, they are basically the same speed, even on a hilly climbing workout.
So, the weight difference of 28 vs 42 is going to be 100g maybe. It's not a lot. Weight vs aero experiments have found conclusions like weight over 100 miles is tiny small compared to any aero gains (talking like +3 lbs having a few minute difference over hours). Which is to say, the effects you describe are pretty small when measured.
Moreover, the choice of 28 vs 42 is rare, feels like is more choosing 25, 28 or 32; or choosing something in low 30s compared to 40s. Meaning, the weight delta is even less in real world choices. What's more, you can go to 650b and have the exact same tire weights with much bigger widths.
Remember that the wider tires are more efficient at lower pressures than narrower tires. However as the pressures go up, the difference becomes smaller. Even without aero gains, the narrower tire will be faster provided its a smooth surface.