I am starting it to give useful information. I think later we need some article about this.
So Jurriaan have show me book "Neo4J in Action" where on the first page there discuss troubles of Relational Model, SQL joins and mySQL in particular to solve GRAPH tasks, when you have Table - RecursiveLink (MM table in mySQL).
I should say this pages contains few mistakes or wrong claims.
Inefficiency of SQL joins
To find all a user’s friends at depth 5, a relational database engine needs to generate the Cartesian product of the t_user_friend table five times. With 50,000 records in the table, the resulting set will have 50,0005 rows (102.4 × 1021), which takes quite a lot of time and computing power to calculate. Then you discard more than 99% to return the just under 1,000 records that you’re interested in!
I think even mySQL do not make join in such way. Valentina DB than more.
This is description of theoretical the most stupid simple algorithm.
In the same time, they show benches of mySQL with HUGE degradation.
And this can be easy true.
But, this is not problem of SQL or Joins or Relational model.
This is just a lack in mySQL of recursive commands.
Oracle have such commands. Valentina DB have such commands also, and even better of Oracle, thanks to our BinaryLinks.
They are describe in our article Hierarchical (Recursive) Queries in Valentina SQL