https://en.m.wikipedia.org/wiki/Floating-gate_MOSFET
Tunnel junctions are used in applications such as photovoltaics as well as high frequency circuits e.g. tunneling diodes.
Edit: Just to fix a confusion, by 'reliable' I meant, one electron sent, one received. I know one 'bit' isn't typically sent as one electron, but being reductive for comparison's sake. Obviously with enough current you can make the tunnel happen, but you'd need far more electrons and 'time' to do so.
The reason we don't often use tunneling diodes today in most devices is that they are expensive to make compared to many other modern diodes, they are still used in more unique applications including very sensitive scientific equipment and space applications.
In general there many types of semiconductors that employ a Quantum Well https://en.wikipedia.org/wiki/Quantum_well this means that these SCs do RELIABLY utilize quantum tunneling by either inducing it under very specific conditions, or averting it.
I'm really not sure what you are skeptical about, without use being able to utilize QT reliably we wouldn't be able to make modern semi-conductors they would either not work because we can't make tunneling happen when we need it too, or won't work because we randomly induce it to happen.
It doesn't mean we fully "understand" Quantum Tunneling as in understand exactly why it happens, but we know enough to predict and control it, Quantum Tunneling is to the SC industry what fire was to humanity in general for milenia. We still probably can only roughly model the physics of fire, especially since at the end it's governed by QM effects, but we don't need that in fact we don't even need to understand combustion that well to utilize it.
But if you want single electron tunneling then yes it's possible there have been papers about it for a few decades with the latest ones also including some experimental work https://arxiv.org/ftp/arxiv/papers/1204/1204.5539.pdf
Single electron transistors are also being actively worked on https://en.wikipedia.org/wiki/Single-electron_transistor
Sounds awesome. And thanks for the explanation.
Just out of interest, I use the extremely fast switching times of tunnel diodes to make quick determinations of oscilloscope rise-times. When used as a switch their rise-times are much faster than most oscilloscopes so the rise-time seen on the oscilloscope is actually that of the scope itself.