Edit: qualify "at all times", "late 80s, early 90s"
[0] http://www.retrobike.co.uk/forum/download/file.php?id=35724
[1] http://www.retrobike.co.uk/forum/download/file.php?id=258229
[2] http://i41.tinypic.com/j9y5ug.jpg
[3] http://fcdn.mtbr.com/attachments/vintage-retro-classic/12540...
Buy a decent frame and build your own bike. I have three bikes -- MTB, hybrid and one I made myself. The homemade one is easily my favourite, and I think it's the better bike. You get to decide what's important to you rather than the manufacturer. Road wheels with bullhorn bars and 3 cogs at the front? No problem, I can do what I like.
If your route has a lot of uphill, then I'd expect that 1.5:1 ratio to get closer to 1:1 where the work you do is more fighting gravity than inertia.
https://en.wikipedia.org/wiki/Rotational_energy
The wheels have a rather large diameter and go pretty quickly. Both diameter and angular velocity are quadratic terms. You could now do the math to see how much energy it takes to accelerate a bike to a given speed.
But you're right, going uphill or at steady speed completely changes the picture.
[0] http://velonews.competitor.com/2012/06/bikes-and-tech/techni...
[1] http://www.tuneruniversity.com/blog/2011/04/part-i-why-wheel...
Aerodynamic resistance becomes greater than rolling resistance very early. See eg. http://www.sheldonbrown.com/rinard/aero/formulas.htm
You may argue that people riding upright on city ("dutch") bikes make up the majority of bike riders, but their possibly heavy-rolling tires are made up for their aerodynamically pessimal riding posture!
(Of course often it's easier to do something about rolling resistance than aerodynamics, despite being a smaller component)
My dual suspension carbon mountain bike is ~23-24lb w/ pedals.
My steel touring bike, well, it's a little hefty.
But there is probably one other factor to consider -- the lighter bike is more expensive, therefore probably has better components (chain/shifters/bearings/tires/gears), which makes it feel smoother. And if it feels better, you will get the perception of higher speed or less effort.
And like you say, better components can make a difference, within reason.
However there is a noticeable difference in riding as well, unless you ride only on short distances.
You can simulate the difference on the same bicycle so the comparison is not tainted by different tires and mechanical components. You can add 5 kg (or 10 pounds) here and there or even in a backpack (but it will harm your performance in other ways). Maybe don't put weight on the steering/front wheel because that will make for a difficult handling. You can use water, so you can release it at any time and feel the difference.
You'll notice that with the extra weight you accelerate less and climb worse. All sort of disadvantages derive from that. If you do 100 km and you have a choice, you'll go for the lighter bicycle. Unfortunately that means a lighter purse as well, maybe much lighter.
It gets a lot more expensive once you go sub 14lbs, and very very expensive sub 10.5lbs or so. (20k+)
http://c.files.bbci.co.uk/16E6F/production/_84270839_froome....
Cycling is unlike any human powered sport except iceskating in that aerodynamics play a huge role. That means packs are the default state of play, except in a few limited circumstances.
One of those cases is climbing, where speeds can be low enough that a few pounds makes a big difference.
When I was towing an empty 20+ lb trailer, and the bike +rider was in the 210lb range (so, 10% of the mass) I was about 10% slower than sans-trailer. Towing a trailer + 25lb kid was about another 10%.
A lighter bike _may_ feel different, but whether it's weight or something else, or even placebo is up for discussion. It's hard to tell frame weights. 20 years ago, I ran some blind testing where we put weighted waterbottles in a frame, one with lead, one with glass beads, and one empty. 5 trials, A/B, which one was heavier. IIRC, adding 5 lbs of lead was noticable to almost everyone, but the difference between 2.5 and 5 lbs was only noticeable to the frame builder.
But we are very bad at judging absolute speed without technical assistance, while being quite sensitive to acceleration. Therefore, a slight increase in acceleration for a given effort (or decrease in effort for a desired acceleration) is very noticeable, while an increase in absolute speed would not. So a lighter bike will feel noticeably faster (even though it isn't), while different aerodynamics (e.g. "faster" clothes that give a lot more improvement in absolute speed than a weight difference, even for weak riders) might give a clear advantage on the clock, but you won't be able to notice it while riding.
Light bikes just feel fast, very much so actually. It's a powerful illusion that only gets stronger because it is supported by those few specific situations (stop and go, races with tight corners, all that climbing road cyclists love) where the weight also matters objectively.
See the 1874 racing bike pictured in this PDF (p. 4): http://library.la84.org/SportsLibrary/JSH/JSH1999/JSH2603/JS...