I am not sure this is true - I actually prefer a heavier bike for general use. If I was racing up hills or still doing cyclocross then light as possible would be best (and of course also me being light as possible being the biggest factor). I have ridden a lot of bikes and the ones that "feel" the best have a moderately "heavy" frame designed to be fairly stiff but not crazy stiff (heavy meaning at most the light end of a steel framed road bike) with extra sturdy rims that are also as light as possible. By far the biggest effects on "feel" however tend to be tires (skinny tires suck) as you can fine tune the pressure to the surface. And of course quality bearings in the BB and hubs can make a huge difference but once you get to a certain middle level they are all good enough.
When the bike gets too light you feel the effect of bumps and terrain much more and it throws the bike around more (especially if you are "getting light" over said bumps). The lightest bikes also tend to be made for racing and can sometimes be on the ultra super stiff end of the spectrum.
I'm not that old, or have a super long history with cycling. But out of the handful of bikes I've ridden, I def prefer double-butted steel (over straight-gauge Hi-Ten, straight-gauge 4130, 90's aluminum, modern aluminum, modern steel gravel bikes) whether I'm going all-out, or cruisin. And even between 2 double-butted steel frames from different tubing manufacturers made for the same use-case more or less, they've felt different enough to me. At least that's how I felt after having put the same components on two steel framesets, and putting them both through about 2k miles each. One frameset was definitely more enjoyable, while the other felt lacking to the point where I wished it were the other bike.
I have two relatives who have raced bikes in the past at high levels, and both owned stores. One of them was a stronger racer who didn't care about gear as much. The other was still a heavy hitter, but wasn't as naturally talented, and enjoyed the nuances between components way more. He and I seem to agree on things.
I do know that tire pressure and width make a huge difference in plushness, but I still prefer a springier ride-feel from the steel I like, vs. even a modern steel gravel bike with 42's with lower PSI. I had such high hopes for that ride, but man, it just felt kinda muted and leaden, and I missed my steel road bike.
But yeah everyone will have their personal prefs, subjective interpretations, and theories about what works for them and doesn't. Gimme that springy lightweight steel any day.
I wonder if that's really true on hills. Most people ride hills with more effort, pushing down on one pedal at a time. The bike speeds up and slows down much more with each pedal stroke. Compare to the flats where you have a smooth stroke and very consistent output.
I could see all the mini accelerations adding up to a much larger energy expenditure than would be indicated by a spherical cow in a vacuum.
Eh, it's really quite small, unless you have stupid heavy wheels (meaning, that the wheels are a significant percentage of the mass of the entire system).
If the mass of the wheel is 100% located at the rim, the energy goes 1/2 into straight line energy and half into rotational. But -- if you slow down, you don't lose that energy, it just redistributes to gravitational potential. The flywheel effect from heavy wheels may actually help, as it tends to smooth out pedal strokes.
The energy to go up a hill is just mgh - friction. Some cadences are less efficient than others (e.g. for me, standing is more power but lower efficiency), and bikes that are too stiff may not help you get the best rhythm.
In a totally unscientific test -- I've got a 12kg Aluminum gravel bike, wide tires, rack, fenders. I've also got an 8kg carbon racing bike, skinny tires, aero rims, etc. They have similar riding positions (reach/drop), though the gravel bike has a wider seat, which is better for one climbing cadence. I usually ride the gravel bike these days, because comfort.
Two weekends in a row, I did a 12x rep climbing workout (2.5 hours), one on each bike. First weekend, Road bike. It felt _fast_. Quick, lively, accelerated from the stoplights on the way to the hill. Second week, back on the gravel bike, grinding up the hill. Total ET difference: 5 seconds.
Rotational inertia matters extremely little for bike feel compared to the overall weight of rider+bike. The torque required to accelerate a wheel by itself without any resistance is minimal.
You have any sources for this? Not trying to call you out but I am generally curious. We know on a car that rotational mass makes a huge difference but a car is continually accelerating and has a constant type of locomotion that is very different than a person riding a bike. In addition the rotational mass of a bicycle wheel is much, much larger diameter comparatively to the total weight of the vehicle (compared to a car) and provides a gyroscopic effect that is critical to the way a bicycle rides and stays upright. I would expect if we could make impossible light bicycle rims and tires the bike would have trouble balancing and feel like shit and also decelerate much faster. But surely there is some middle ground that is ideal?
It's first year kinematics. The (first order) math is pretty easy to work out.
1/2 I w^2 is the rotational potential energy, Assuming all the mass of the wheel is at the rim (worst case) I = mr^2, w=v/r, so that works out to 1/2 mv^2, which is the same as the translational kinetic energy. So the 'energy penalty' of rotating mass is 2x. So, a 10kg bike with 2kg of wheels and 100kg of rider is approximately equal a non-rotating mass of 112kg. However, that rotating mass gives you 2% more kinetic energy to trade for wind drag, gravitational potential, or other friction. So, not much.
A bike's does not require gyroscopic effects to stay up -- there was an experiment with counter rotating wheels (to cancel the gyroscopic moment) that was ridable. The actual stability depends on the geometric and pneumatic trail, flop, weight distribution on the steering axis, and a few other things.
There is a middle ground that people have converged to -- Somewhere in the range of around 600mm bead seat diameter and an inch or two of tire. (26"->700c, 25->50mm width covers most of the biking world.)
You can test this yourself. Flip your bike upside down, shift it into a gear that you use to accelerate, take a torque wrench with hex socket and put it in the left pedal socket on the backside of the crank like you would when installing it. Then test the torque required to get the wheel moving. You would find that these numbers are minimal.
Also gyroscopic effect has very little to do with the bike staying upright. The reason a bike stays upright is because of the trail interaction and camber thrust. When the bike leans left, camber thrust of the tire makes it turn left, which moves bottom of the bike to the left, correcting the initial lean, while the geometric trail prevents the tire from turning left completely.
I've ridden heavy bikes, and I've ridden light bikes. Light bikes feel better, rotational weight mattering more than anything else.