A big problem with pricing rare records is getting enough data points for any one album. Yes, in theory you may find 1-2 rich collectors that will pay a huge amount for something, but in theory there's someone who will pay me $1,000 for my dirty socks- that doesn't reflect the real market value of my socks

I might have to wait 50 years to find that $1,000 offer- if I want to sell my socks in a reasonable amount of time, I ain't gonna get that much for 'em. We see this on ebay all the time, with people doing 28-day listings with outrageous prices listed month after month, hoping for that one 'special' buyer to come along...
For a more metal example, ok, someone paid $3,000 for Flames of Hell... if you had 5 copies to sell, could you find five more people willing to pay $3,000 each for a copy? If so, ok, you've established that the album will REPEATEDLY sell for that amount. If, however, you sell 1 copy for $3,000 but no one else will pay more than $1,000 for one of the other four copies, what's the item's value? IMO it's $1,000 and you just got lucky on that $3,000 copy.
Records like FoH often bring huge amounts simply because there are not enough data points to define a value, so people pay astronomical sums based on the item's apparent rarity. I say 'apparent' because more copies could always turn up. The more copies that turn up, the lower the price drops and the better the chances that a 'regular' price can be established.
Example- first time I saw a Dutchess 7" on ebay, it sold for 411 GBP (over $600). TMK it was the first copy on ebay, and no one else had offered a copy for sale in years, so it went for a big amount. Since then, around 20 copies have come and gone on ebay alone, and the price quickly dropped and became consistent around 150-175 GBP ($225-250). Every now and then a copy has sold for noticeably more or less than that, but there are now enough data points to get a feel for what most people are willing to pay for the item.
Some important points:
1- if you're buying based on claims that only X copies exist, unless that's a manufacturing number, more copies could always turn up! Yes, only 2-3 Warzwolf LPs are known to be in collectors' hands, but I guarantee you that more than 3 copies were pressed, so more DO exist (we just can't find the darn things!

). On the other hand, only 10-20 yellow Avatars were pressed, so there really are only 10-20 of those out there.
Another example- I was talking with a blues collector years ago and asked him about a Gatemouth Moore LP that was listed in a price guide with a note, "only 10 copies known to exist". The guy walked over to a shelf, pulled the album out to show me, and said, "Yeah, well, better change that to eleven... the author never asked me if I had a copy"
2- If enough copies have sold in recent years, what is the AVERAGE selling price? Popsike is great for this (not foolproof, but very helpful).
3- Check the bid history and se how far back the 3rd, 4th, and 5th place bidders were. Example, that Bad Axe LP that sold on ebay last month. 2 bidders pushed it to $686, but 3rd place was around $350 and 4th and 5th was a lot less than that. The album has never brought that much on other occasions, so this was an anomalous bid war, not the typical value of the record.
4- What is the item worth TO YOU? Things only have the value we give them. If Flames of Hell is worth $3,000 to you to own a copy and you have the money, then it's worth $3,000 to you; it is not worth $3,000 to me, but that doesn't really impact how much you value the record.