They did a 'sell as you go' calculation as well, look about half way down the article.
The real flaw in their approach (in my opinion) is considering the used mining GPUs as being worth nothing when calculating net gain/loss.
I do note that you would still have the hardware, which is why a lot of mining firms are trying to sell off their hardware -- to get out of heavy debt.
The problem with trying to do every potential calculation is that it just starts bloating the article, and there will always be some other piece of information you could add.
I didn't include explicit space rental, PC cost, AC cost, or numerous other factors in the calculations. I only guesstimated at GPU costs, because they were all over the place.
The main point is that if you used an RTX 3090, 3080, 3070, RX 6900 XT, etc. for mining, hoping to get a bunch of coins worth a lot of money, here's what you would end up with:
2.7 ETH (max) from 24/7 mining with a 3090
2.3 ETH from a 3080
1.0-1.2 ETH from a 3070, 3060 Ti, 6900 XT, 6800 XT, or 6800
When people see pictures of mining farms that had hundreds of 3080 cards or whatever, and think "Wow, just imagine all the cryptocurrency they mined!" They're forgetting the cost of all that hardware and ultimately the bottom line. If they mined ETH and sold at the right time, yes, they made money. But these companies got into this for greed, and likely failed for exactly the same reason. "It will go back up, just wait and HODL!"
2.7 ETH total from a 3090 that might have cost $2500 or more when purchased, minus the power and system costs, basically ends up at close to a wash. That's assuming the card still works. Sell it off and you might get $1200 for the GPU now, which means in total:
$2970 from the ETH at current value
-$500 from electricity costs
$1200 from the GPU being sold
Minus other expenses not factored in -- how much time did it take to set up, monitor, etc. the mining equipment? What about AC costs? What about the rest of the PC? Was this a business where there were other employees?
Most miners, even small-time miners, probably spent hours every week of mining ensuring that the systems were up and running and performing as expected. Sure, it was sort of a "hobby" so that was free time, but let me put it this way. If you told me you would pay me $3000 after 20 months, and all I had to do was check on a few things every day that wouldn't take more than 10-15 minutes, that's 150 hours of potential work. $20 per hour might seem like a good income to some people, but it's a lot less than any IT professional capable of running a mining farm would make doing any reasonable job.
Okay, that's not entirely accurate. Let's say a small-time miner had two full rigs of six RTX 3080 GPUs that were originally purchased at $1500 each (being generous). Call the rest of the PC cost $1000. That's $20,000 in hardware expenses at the start. Then they mined 24/7 for 650 days and brought in 24 ETH total (I'm going with a more conservative estimate of ETH mined, due to potential downtime, pool fees, etc.) Power cost at $0.10 per kWh would be around $5000 (higher than the estimate put in the charts, because of extra cooling fans, PSU inefficiency, and the rest of the PC). Today those RTX 3080 cards sold on eBay might get $700 each. The rest of the PC hardware might be worth another $1000. So, $20,000 in initial expenses, $5,000 in power, and you get $26,400 from the ETH and recover $9,400 from selling the hardware. That's $10,800 net gain.
It's certainly something, and then you scale that up to four or eight mining PCs. More than that really isn't feasible without setting up a dedicated warehouse IMO, because eight mining rigs each pulling down 1500W–1800W of power is going to be about the limit for most homes. Monitoring and maintaining eight rigs is also still something that could be done with only a few hours per week on average. But 12,000W–14,400W of power being consumed
constantly inside just about any home would make for a very toasty living space, even in the winter. (I know a guy that basically did this in 2017 and he didn't turn on his heater all winter long, and still had a bunch of windows open in the basement with fans blowing in cold air from outside. That worked okay in the cold, but summer was a problem.) You'd have an electricity bill just for the mining rigs of more than $1,000 per month.
Also fun fact: I have been offered consultation jobs multiple times in the past two years to help companies set up mining farms. I turned them down, because I did not want to get involved with what would very likely end up as a losing operation. Been there, done that (in 2014).
Here's another interesting question: If a company bought a bunch of hardware and mined ETH, and was heavily in debt for the hardware, and then declared bankruptcy... would the ETH have to be sold off first? I suspect the heads of companies are syphoning off some of the ETH or BTC or whatever just in case this happens. "Sorry, our endeavor went bankrupt. Here's everything!" [Doesn't mention that 1% of the miners were actually sending ETH to a different wallet.] The more I look into cryptocurrencies, the more I realize the whole operation is entirely shady and questionable. I get why lots of people might value that sort of thing, but its inherently shady nature and immutable transactions means it can never go mainstream.