AJs is within a 15 percent opening range in equitylab.
55 is only within top 26 percent of opening hands.

So 55 is ranked lower than AJs.

But 55 is a 51.32 % favourite against AJs.

When 55 is a favourite to win then how can AJs be ranked higher than 55. This makes zero sense to me. 55 is clearly a better starting hand than AJs.

Playability should not matter in rankings because equitylab assumes that you are all in.

Try playing 55 and AJs against a 10%, 15%, or 20% range. You'll consistently see that AJs does better than 55. Sure, if you have to choose between one and your opponent gets the other, 55 would be better. But, your opponent is going to play a range of hands. And, 55 just doesn't do well against reasonable ranges. Heck, even against a completely random range, AJs does better than 55.

AJs does poorly when you are up against pairs. But, it does well against ranges that are heavy in weak aces, and unpaired broadway cards. Those same ranges are usually a flip against 55 or have 55 crushed (the bigger pair portion). Consider a hand like 99. Which would you prefer, 55 or AJs?

As for playability, that is a whole different situation.

Edit: Just for a note, AJs is within the 4.3% range in Equilab. So, it's not even at the bottom of the 15% range.

Okay. So I guess equitylab compares how a hand will do against a random hand in the rankings.
I compared AJs and 55 vs a 15.08 % opening range and it does look like AJs is better. But very few hands in a 15.08 % opening range is better than 55. 48 combos out of 200. 80 out of 200 combos in that range is better than AJs. I guess AJs is better because it has more equity against many hands in a 15.08 % opening range, and many of those many combos that 55 is better than it is basically close to a coinflip so not a lot of euity there contributing to the overall score in the range vs range calculation.

As Vinnie says, you want to think about how hands do against ranges. I would much rather have AK than 22 even though the low pair beats the two offsuit high cards if they're all-in together.

There's another problem too with the idea that whatever wins in a heads-up battle should rank better.

AK > T9s ==> AK should rank better than T9s

T9s > 22 ==> T9s should rank better than 22

And yet...

22 > AK ==> 22 should actually rank better than AK

So there's no easy way to get rankings that don't contradict each other purely based on what hands beat others heads-up.

I don't know how they got their rankings. It may have been by running each hand against a random hand. I tried to look into it, but I wasn't able to quickly find something.

It isn't about how many combinations of hands 55 is better than.

48 out of 200 combos are better than 55. This is true. But those 48 combos are 77-AA. And they are an 80.8% favorite against 55. The rest of the combos {A7s+,K9s+,QTs+,JTs,ATo+,KTo+,QJo} have 47.7% equity. When you hold 55, you're either crushed or slightly ahead.

That same range is a little different when you hold AJs. First, there are fewer combinations. You are crushed by AA (12.9%) but that is only 3 combinations not 6. You're in bad shape against JJ-KK (32.6%) but that is only 15 combinations not 18. We are in bad shape against AQ+ (30%) but again we block a lot of those combinations (24/36 remain). In fact, there are only 163/200 combinations left. Against 77-TT, we are 47.1%. The rest of the range, though, we are a 66.2% favorite against. That's like 101 combinations (out of 163) that we win 2/3 of the time.

55 never has a lot of equity when it is ahead. It is absolutely crushed when it is behind.

AJs is rarely crushed {AA}, usually has more equity when it is behind, and often is ahead as a 66% favorite.