Not really. I honestly don't care about address puzzles, because of this reason:

There is no incremental progress to speak of. Every time you hash a key, you have the exact same chances of success as before, even if you scanned all keys except the very last two, or if you're just trying for the very first time some random key.

People who don't agree with this are ignorant: they ignore the fact that there was effort already performed. The only thing to track is the scanned regions, but this is not exactly progress. Progress implies that somehow something's better, which is not the case here. It's just economy: you trade time/resources/money with the chance that you might win, or you might lose everything. However, the chances are 4 to 1 not in your favor (every time you pay for hashing a key, you "win" a quarter of your expenses; and that only ever happens after you played millions of dollars). Not a good gamble, lottery is way more profitable if you play it smart (though it's still a losing gamble).

If someone discovers a way that the effort actually improves on anything, then we'd need to rewrite most math books. That's why it's really easy to call BS instantly on some people around here...

lol yeah. it's just a little above 1 2**70 for every time you do a H160 and check for a match.

"A little above" meaning insignificantly close to 1, since there's the additional (almost zero) chance that, besides the creator's private key, there might be other keys hashing to the same address. We can't know that, but if does affect the chances (as minimal as it does).

It's also 100% if you ultimately intend to scan all 2**70 keys (assuming the creator isn't trolling).

Of course they will get solved, it's just a matter whether the risk is worth it or not. If BTC price goes to the moon it's obviously worth it, as it was for 67, 68; 69 was too risky, and also kinda bad luck due to the 0.7% offset. But 71 isn't worth it today if one goes serious with it (instead of crazy).

You should not forget that your PC's power consumption heavily depends on the task it's doing. It can vary a lot from a few Watts to several hundreds.

When you crunch 24/7 using 3060, that's a way more than 270 Watts per day.
Just the 3060 alone draws about 150-200 Watts, that's not including CPU, RAM, Mobo, Monitor, etc. If you use 3060 for 24 hours (a day), that's like 4000 Watts*h per day without including other components. 15 times more than "270 Watts per day".
If you don't pay for your electricity or have a flat fee, that's different.

Got it. I guess you don't load your PC all the way up if you pay between 80 and 150 only. If you do, that's like one more AC running all the time.

You mean when they average monthly payments based on your full year's consumption? Yes, some of them do. In this case it's harder to track what you use even though you pay the full price anyway, right.

85 seconds for puzzle 48, equivals to aproximately 45 years for puzzle 71 at, think about it

It has been debated to death (and proved in all shapes and forms): "luck" is not affected if you scan either sequentially vs. some picked untested key (either at a random position, or via whatever other made-up logic, like prefixes).

What gets affected though, is the efficiency of the computations, and there is absolutely nothing currently that is known to be more efficient (e.g. that costs less or runs faster) than a full-on sequential scan. That is, if one wants to solve anything in the most efficient manner possible, not playing bingo with the elliptic curve and two uniform hash functions.