The argument that you can get your hand knocked off the hoods easier isn't really valid anymore, in my mind. Maybe back in the early days when hoods were tiny and position deeper onto the curve part of the drops. But now pistol-grip hoods are a pretty rock-solid position.
As someone whose stable includes a couple bikes with old-school non-aero brake levers, I'd actually argue that security is a bigger issue with the new stuff. Modern pistol-grip hoods don't have any more vertical hand retention than narrow old hoods, and unlike the old hoods which are best thought of as grip nubs, the new hoods go farther in encouraging you to use them as loose handrests. The danger is when you're draping most of your fingers lightly on the brake lever and you hit a big bump/pothole/whatever.
Neither style is very good at simultaneously providing good brake access and a secure grip from the hoods. If you fold your lower fingers below a modern hood while resting the index finger on the brake lever, as people often do, the lower fingers usually interfere with lever actuation and the index finger doesn't have good leverage on the lever.
Not that I think it's a huge deal either way... I think most people make pretty good judgement on where their hands need to be at a given moment, by feel.
Another part of the puzzle, to why people didn't spend more time on the hoods back in the day, is the position of the brake lever pivot. Old-school road levers are very similar to flat-bar levers: the cable exit is roughly inline with the visible part of the brake lever, and in order for the lever to actually pull cable in this configuration, the lever's pivot needs to be offset from the lever:
If you're on the hoods, this means that the lever's pivot is between your palm and your fingers. So, in order to actuate the lever, you need to "push down" on it with your fingers as much as squeeze.
The resulting cable routing is excellent at minimizing friction, and it's kind of neat that you can maintenance brake cables and housing without tearing up the bar tape. But, that "push down" operation makes for inferior braking from the hoods, both in terms of power and control.
Contrast with a modern lever, where the pivot sits right at the top of the visible part of the lever:
The direction that your fingers need to pull on the lever to apply torque is much more toward the palm, and that becomes even more true as the lever is actuated farther, so braking force is controlled through a squeezing motion of the hand... significantly more effective.
As for the drops...
The non-aero brake levers pictured above are on my 1983 Miyata 710. I purchased it a couple years ago to stand as an fun contrast to my Emonda; in its day, it played a similar role in its market as my Emonda did when I bought it in 2015. They both have standard recreational 2x drivetrains for their respective era, the steering geometry is very similar, and most importantly, they both use vibrant red color schemes.
Anyway, aside from the one really big modernization of replacing toe clips with SPD-SL, I've kept the Miyata basically vintageish. And that includes bar tape: when it came time to re-wrap, I replaced the original cotton cloth with new cotton cloth.
And for funsies, I decided to take the extra not-necessarily-period-correct step of shellacing the tape. So the final bar wrap is only about 26mm in diameter, and it's rock hard.
What's my point with all this? Well, basically, what surprised me about the old Miyata is that my hands are totally comfortable on it. I've done century rides no problem on it, and like all my other bikes, I never ride it with padded gloves. According to everything I read, having these tiny little hard cylinders boring into my hands should be immensely jarring. But it's fine.
Or at least, it's fine when the angles are right. A fun little feature of the bike's quill stem is that SR did a crap job on the bar clamp. I've had it slightly slip on me a couple times while riding over bumps, and then the drops become quite uncomfortable
until I rectify the fit.
The handlebars installed on the bike currently are a Nitto B115 Olympiade, which are more or less a classic Maes-bend bar. They've got a fairly round bend - making for less extreme ramps than some vintage bars - and a little bit of flare.
One thing I notice is that my forearms never interfere with the tops when I'm in the drops. This seems to be an issue with some modern bars, due to combination of zero flare and a long reach. Some bars solve it by using a shorter reach and a longer stem. Alternately, flare obviously helps; it sounds like a bad idea for road bars, but a small
amount of it doesn't seem to diminish the road attitude of a handlebar.
But the other thing is, the curve of the bars seems to mesh extremely well with the channel of my palm, from the hooks down most of the way through the drops. I think I actually cock my wrists less in the drops than on other bikes, because the good mechanical conformance between my palm and the narrow bar seems to help get a secure connection, without my needing to put my weight on the inside and cocking the hand aggressively to the outside to form a big platform.
This isn't to say that I'm going to change out the gel cork on my other bikes to shellac'd cotton cloth, but I do find it interesting. What I wonder is if a lot of drop comfort issues aren't simply due to the overall shape and position of the drops receiving insufficient attention, rather than any need for exotic non-cylindrical drop shaping. Even if there's a real need for non-cylindrical drop shaping, it's pretty easy to accomplish with the finishing work rather than with the underlying bar. I suppose doing the re-shape through carbon layup can save a few grams, but in my book, that's nowhere near being worth yet another bicycle industry standards catastrophe.
From the article:
Either one system is wrong or the other is.
No. One system is simply less-adjustable than the other. When you couple parts and reduce adjustability, it doesn't make the part right or wrong, it just limits its usage scope (usually in exchange for some structural benefit (i.e reduced weight).
When things get rigidly coupled, far more part permutations are required to achieve a comparable adjustment scope as before, and the end user has less opportunity to change things without further cost. The big question is how much adjustability is merited.