I'm going to keep posting updates here in case someone else who is in the same situation finds it.
It seems like putting a coat of Krylon Fusion on 95% of the concrete did not fix the problem, it just delayed it. Or something. My pH still spikes, it's just not consistent in when or how much.
It also took ages for the spray paint to dry. I left the tank to dry about 48 hours. It still smelled like paint when I filled it, but I assumed it was just the tank that wasn't getting aired out well enough. (Note to other people doing this: the tank won't smell once the paint is dry). The tank smelled like chlorine from the uncured paint, and I ended up emptying the tank and leaving it to dry five more days. It stopped smelling on day 3/5.
I filled the tank and left it. The pH started at 6.4, and 36 hours later it was at 6.9. 48 hours after the initial filling, it was way higher. I didn't actually bother to use the high pH test, but it was at least 8.2. I suspect either there were lots of bubbles on the concrete preventing the water from reacting with it, or the water needed time to soak in and reach the coated concrete.
I might have discovered what was different between the tutorial's concrete and mine, however. On closer inspection, the only difference between the two tubs is that the tutorial says "Contains portland cement" and mine says "Contains portland cement and lime." And lime is known for raising pH.
Now that I have a likely cause, there's something I think I can try. According to the wikipedia page on
Calcium Hydroxide (slaked lime), Ca(OH)2 dissolves in water, raising the pH to up to 12.4. Ca(OH)2 may then react with CO2 to produce
Calcium Carbonate. Calcium Carbonate is not soluble, unless it reacts with more CO2 to form the soluble
Calcium Bicarbonate. As far as I can tell, Calcium Bicarbonate does not change pH and is not harmful to fish? In any case, you can remove it from the water by doing a water change.
So would adding CO2 to the tank speed up the stabilization? I've been soaking the background without a filter running because that requires half the amount of water. Maybe that was also a contributing factor since it significantly reduce air exchange? Either way, I think I'm going to try adding CO2. There is a risk that the calcium carbonate will form on the cement itself and turn the whole structure white, but I haven't really got anything to lose.
I was intending to set up a CO2 system and make this a planted tank at some point anyway, so I went ahead and bought the supplies that Joey suggested in his
DIY CO2 system video. (If it doesn't end up being useful in this tank, I'm sure I'll end up using it somewhere else.) It cost me about $60, but if you only wanted it to stabilize your tank, not as a long term planted tank solution, you could probably manage $30 or less.
These supplies arrive on Monday. I'll set it up at some point and post another update then.