There are some problems that can be solved by quantum computing much faster than with conventional computers, but there are many types for which quantum computing is no greater advantage at all.
Along with scaling practically indefinitely in price and power consumption.
Yes, at least in real-world, practical senses.
Computer logic circuits are two-dimensional, and the output from one flows directly to the input to a very small potential number of others. Neurons are arranged in 3D and can directly connect to thousands of other neurons. Emulating that very aggressively 3D arrangement in 2D computer chips will use vast amounts of chip surface area. In an industry where chip areas are measured in the 10s or low 100s of square millimeters, it might take square meters of chip to fully simulate a human brain.
My mentions of Moore’s law was to point out that we are almost at the very, very, very end of how small we can shrink components on a silicon chip–we are at or almost at the point where the placement of individual atoms makes or breaks the function of the chip. And they are already very leaky due to quantum tunneling, so that you have to pour more electricity into them. You know what you call a very thin wire with lots of electricity being poured into it? An incandescent light bulb. And the faster you switch the chip on and off, the more heat is going to produce. So you will need more and more aggressive cooling technologies.
Speaking of how fast you switch the chip on and off, at the speed modern computers operate, even light moves slowly. Even if you had optical links between chips running at full clock speed, with a 4 GHz system (for example) the signal would more less than 8 centimeters. So there is a limit to how fast information can propagate between the many, many, many chips you would need to run your human brain simulation.
So, let’s say that there is a little more life left in semiconductor technology before the quantum tunneling and heat problems become utter brick walls. Let’s be very generous and say that these future advances will mean that simulating a human brain will take a minuscule 350 thousand CPUs instead of 350 million in my earlier mention. Let’s say that you can get an excellent bulk price of 50 bucks each for those CPUs. That’ll rack up to only $17,500,000 for the CPUs needed to build your human brain simulator. That isn’t counting the extra computer hardware, the pumping for the liquid nitrogen to cool the chips, or the dedicated nuclear power plant to run the computer. For one person.
Uploads are not going to happen. They are just one more incredibly silly science fiction fantasy, like cheap casual interstellar travel, antigravity, hologram projectors, and time machines.