Retired in 1997 after 31 years in a high-tech factory. No problem on any one question, but maybe getting old, "GPIO" doesn't ring any bells. Maybe we used other terms. Or, maybe I knew it when I retired, and it has fallen off the memory banks.
We kept old equipment running for years and years, sort of a skunk works operation because of our skilled people. If we had a problem on an old HP9825 desktop used for test equipment on high production products for military aircraft, the engineers would laugh at them when they asked for help. They would come and wake me gently (I was not an engineer, just a production tech) and I would have to sort it out, and write them a users manual on how it worked. Took me two weeks if all documents were lost, but new equipment would have cost hundreds of thousands of dollars, with design and programming costs. The engineers could only do PC stuff, or standard programming languages. I was more of a generalist. They could do Fortran or ADA or C++, but to decipher an in-house interpreter language for a 9825 with all documents missing, they were lost.
I was a high level technician, but I would have to pull out the ARRL handbook before I actually could design a low pass filter depending upon frequency and power levels. But, one would think they could draw you a rough schematic with no values stated. In fact, on the job test for our labor grades, there were a number of questions exactly like that. Low pass; high pass; band pass, Chebychev, etc. And, you had to know without values, by where the components were located. For example, two caps to ground with a resistor between them will shunt all high freq. stuff to ground, etc. You know that without values stated, you just don't know the frequencies involved.
I assume 0.65 v forward biased, right? That is, standard silicon diode forward bias with some sort of current limiting resistance. Unless they are cascading them now. It has been a while and I am sure things have changed a lot in 14 years.
I did know some men who really knew their stuff. We had a power supply on a Boeing cockpit display unit, once, that was burning up coils used in inductive output filters. They brought back a retired engineer, and I listened as he calculated in his head how many turns of what size wire, and how many milliohms the coil was, for dc resistance. 20 milliohms, I think he said, and I had just measured that exact dc resistance of the wire. Turns out the dummies who supplied the coils had changed the varnish to one which melted at normal operating temperatures, and the coils would go massive shorts across themselves when hot, so inductance was very low and the power supply current limiter would shut it off. At the high switching current levels, that 20milliohms was enough to heat it very fast.
I have a friend, a very successful businessman in Puebla (large city SE of Mexico City). He runs a machine shop for high precision products for the bottling industry, at great levels of income for himself. Computerized design, numerically controlled machines. Fantastic operation.
He got in a bunch of interns from machinist school, allegedly finished their academics and math stuff.
One day he asked them how many milisimas (spelling, not sure) which in Spanish means thousandths, in an inch. The biggest number guessed was 50. He realized he had a big problem, and started from scratch with them.
Frankly, the sort of problems you list could be done by our tech school graduates with a couple years experience, and to this day they make a lot less then $60,000 a year.