In the context of the point he was making, they certainly are.
No, they are not - far from it.
[t]he end result is still the same with either. A faster way of transport.
There's hardly a change imaginable as massive as the step from horse to car.
A horse can transport up to 2 people - though one is better - and a little freight. A car can scale very flexibly to adjust to the owner's needs and can transport up to 5 people easily (up to 100 if you count in buses) and take freight up to tons of weight.
A car protects you from weather conditions, a horse - not.
A horse can go nearly everywhere, whereas a car needs roads or at least a halfway decent path, so the terrain of the whole world has changed to adapt to the needs of a car. As a result roads changed the importance of whole cities depending on whether the traffic was routed towards or away. Think of Route-66: Since most traffic has been re-routed via the new freeways, most towns along Route-66 have faded into oblivion.
A horse needs food and water in quantities that are hard to take with you for longer journeys, so you need to plan your trips well. A car can take all "food" it needs with it easily without even impairing its load capacity - and can go up to 1200 kilometers without requiring a rest or time for regeneration. As a driver you can go hundreds of kilometers and still feel relaxed upon leaving the car - try that with a horse...
To cut it short - a car is much more than just a faster way of transport. It offers so much more options and higher flexibility that it is not even in the same league with horses.
The horse had reached its maximum potential as a tool of transport, so the next step with automobiles was inevitable.
That's exactly the point - the concept of a horse being the main transport medium had reached its limits, but people would not be able to think out of the box! They would not be able to realize that they need to take a completely different route and leave horses behind, if they want to improve further.
In hindsight the next step with automobiles looks inevitable, but back then it was not perceived that way - at least by the majority. Instead horses were considered reliable, familiar and sufficient. Perhaps you could breed them a little faster even, with better food during upbringing, tweaking some bits here and there... but switching to a new technology that is loud, strange and slower than the trains we already adopted to over the recent decades? No way - instead a slightly improved yet-another-Xeon-box ...errrr..... a new, fresh young horse would be sufficient, wouldn't it?
Humans tend to stick with what they have and know. They rather seldom recognize what a change might bring and so they would not want something new, until it's available and in use by early adopters - only THEN they suddenly realize they indeed want it, badly! Thus it's questionable if it's the right strategy to _only_ listen to the customers without interpreting.
What Mac Pro users want is not a big box w/ Intel server hardware. That's only what they know. What they _really_ want is a machine with lots of power for a fair price/performance ratio, which is reliable, comparably silent under load, offering easy access and some headroom for upgrades and a couple of storage units. Ideally neatly wrapped into a high-quality package in terms of look and feel.
Ignoring the outdatedness of the current Mac Pro, there are only a few limited configurations possible in terms of CPU('s).
Chances are that you either pay more than you would need, if you e.g. like to have the headroom and easy access for upgrades, but would not need server-grade components.
Or you reach the upper limits of what's possible, but can't get more as the maximum is 12/24 cores running at xy GHz. You can buy another box, but you don't have the additional computing power available instantly. You would need to e.g. split up work to make use of the second box you purchased.
With a modular concept done right you could configure the machine in much finer granularity according to your real needs, so you could eventually reach a better individual price/performance ratio.
Make the entry with consumer-grade components (like e.g. a core-i5/i7) or a single Xeon and upgrade later with additional CPU modules.
Perhaps you could even rent an expansion module for that one high-volume project with tight deadlines, which would not justify buying a new machine and migrating everything to it.
Need much more computing power than those measly Dual-Xeon setup? Get a multi-CPU expansion with 128 ARM cores - the supercomputer cluster for the rest of us.
I would prefer a storage expansion module over a rat's nest of cables and bigger footprint for additional external drives any day.
The PSU would not have to be dimensioned for the maximum expansion possible in the big box (and thus be over-dimensioned for 90% of all users), but can be smaller and thus more power-efficient. If you need or want more later on, simply get a suitable PSU expansion module (unless each expansion module has its own, optimized PSU in the first place, hooked to the common power rail).
And so on...
I don't doubt current Mac Pro users would prefer a simple update of the PCB and CPU's inside the existing box, as they are afraid that any change may be for the worse. They know what they got, but not what they may get - just like the people that experienced the transition from horses to cars as primary means of (mass) transportation...
----------
You're completely missing my point...
The term "horsepower" had been invented during the upcoming of steam machines. It helped dimensioning a machine for running a mill or loom and smoothly introducing the new technology to people by using familiar terms and units.
So your point is? A steam machine is a horse, too?