The graph styles are honestly of bad choice and hard to read on top of being odd at times, I don't know how reddit so yes, there links to the images and not the images directly.

Graph one and two: https://imgtr.ee/i/Gmw0I, https://imgtr.ee/i/Gmv3R the red lines are separation between attractiveness, a better separation per see between 0-1, 2-3, 4-5. Blue lines are division between looks level and the green line is the intersection.

Why? Because while we as humans can definitely difference between a person visible ugly and a visible good looking (a 0 is aways a 0 and a 5 is aways a 5) peoples in the middle are more volatile and adept to change based on personal interests ( a 3 can be a 2 and a 3 can be a 4).

The conclusions on graph one being:

A- bellow average women have it very hard when compared with a average looking women.

B- average looking and a good looking women have basically the same amount of messages.

The conclusions on graph two being:

A- women rated 49% of men as ugly, 36% as average and 6% of guys as good looking.

B- while women message guys they do not find attractive the intersection between mensagens and attractiveness happens at 1 and it goes downhill.

Final considerations

We don't know how they reached this message stats but if they considered answers on the numbers it make sense, as if you're top of looks you are going to be seen as a catch and in order to get a date the number of messages required would be smaller when compared with a men in the unattractive side.

This can not be confirmed but presumed by graph three and because men are the ones expected to send and keep the talk going. And you have the male graph that indicates that men rarely goes to 0-1 what is consistent with the male mindset of effort vs reward.

Graph three: https://imgtr.ee/i/replyearned.Gma2B

It's where the theory of few messages equals more dates come from. Male senders have a spike of replies at 5 in attractiveness what basically could not disprove but exemplify how the premise of women messaged more lower looks guys is not necessary a positive thing.

You can also see that the curve on the male end slows down what again confirm the consistency between this graph and the graph of male messages/looks

Graph four: https://imgtr.ee/i/messagereceived.Gml5Q

This graph is kinda tricky, not only because it use the lowest dominator as a basis but also there's the fact that men are the one expected to start.

I cannot offer a good study of the graph. Not only 0-1 women are not messaged what inflate the numbers of the 2-3-4-5 women and such a graph would only indicate for both genders a chance of someone to try and start a conversation, what again is compatible with the male expected to start.