There isn't an objectively correct basket of goods that is obviously the correct deflator to use, which seems to be a significant underlying assumption of your line of reasoning. The debate about which ones are the important ones to include in statistics like the US BLS CPI is a politically charged debate resolved in part by political means, not purely by the disinterested pursuit of truth. The CPI in particular eminently susceptible to drift over the years, since the goods in it change over time according to the Consumer Expenditure Survey — in 1973 people in the US were buying washing machines that are still in use today and Saran Wrap made of actual Saran, for example, and today they're buying washing machines that wear out in five years and Saran-free Saran Wrap. And of course it measures the prices of only mass-market consumer goods, not services (such as the essentials, child care and elder care) or custom or unique goods such as hand-tailored suits or buildings — a political decision, not an objective one.
If we want to be skeptical of carefully tailored metrics with thousands of parameters produced by political appointees, what standard should we use to measure the value of the dollar? Precious metals have been the standard against which currencies have been measured for several thousand years — the gold standard for measuring the value of currencies, you could say — and by that standard the dollar's loss of value since 1973 is about a factor of 25. This compares to about a factor of 2 over the previous 40 years, since 1933, and a factor of about 1.1 over the previous 140 or so years since the dollar was introduced.
I suspect that if you compare other goods which are, like gold and crude oil, verifiably produced to the same standard of quality in 1973 and today, you will find a similar factor of 16–32 in their dollar prices. I'm thinking of the most common grades of steel, aluminum, brass, portland cement, window glass, industrial electric motors, and so on. There will definitely be some exceptions — ±1% resistors are much cheaper now, to the point where you can't even get the ±20% kind that were the norm in 1973, and I imagine the same is true of specialty steels, synthetic sapphire, and a number of other things that were barely feasible at the time; and presumably photographic film has become more expensive, as it has ceased to be a mass-market item. If you're right about the cost of electricity "staying the same" — by which I assume you mean that, in the US, it increases in line with the BLS CPI? — this suggests that my hypothesis won't be true of coal. Do you have any other ideas?
Let's take anthracite, because it's the purest grade of coal, so it should be less vulnerable to variation in value from drift in grading standards. https://www.eia.gov/totalenergy/data/annual/showtext.php?t=p... suggests that nominal anthracite coal prices have risen from US$13.65 per short ton in 1973 to US$70.99 in 2011; https://www.eia.gov/energyexplained/coal/prices-and-outlook.... says that in 2017 they were US$93.17 per short ton, FOB the mine. That's a factor of 6.8, which is a lot closer to 6 than to 25.