Matching multiple independent (or dependent) paths - neo4j

Question

I'm looking at a hierarchy of people and organizations, and trying to find where/if they meet and share management. Let's say "Bob" and "Susan" work for different branches. I want to show their two reporting relationships up through the company if/when they overlap.
This query currently works great, and returns a single path:
MATCH path=(p:Person {name: "Bob"})-[:reports_to*]->(o:Organization {code: "TopOfCompany"})
RETURN path;
This query also works great, and returns a single path:
MATCH path2=(p:Person {name: "Susan"})-[:reports_to*]->(o2:Organization {code: "TopOfCompany"})
RETURN path2;
This query (doing both of them in one operation) returns nothing at all:
MATCH path=(p:Person {name: "Bob"})-[:reports_to*]->(o:Organization {code: "TopOfCompany"}),
path2=(p:Person {name: "Susan"})-[:reports_to*]->(o2:Organization {code: "TopOfCompany"})
RETURN path,path2;
The same is true if I reuse the first o binding in the second path query.
I'm aware that I could reformulate this to find where the two people meet in the middle, like this:
MATCH path=(p1:Person {name: "Bob"})-[:reports_to*]->(o:Organization)<-[:reports_to*]-(p2:Person {name: "Susan"})
RETURN path;
And indeed that query runs fine - but if they don't meet in the middle, this query will fail since the o:Organization in the middle doesn't exist.
There are probably other equivalent ways I could reformulate to get to the right results - but the heart of my question is, is it not possible to identify two different independent paths in one query? This would be useful in the case where they don't meet, where the targets ("TopOfCompany") I'm matching to are different, or I just wanted to compare a series of paths.
Oh, and I'm on 2.2M04, using the server. The query with two paths succeeds, but the results are empty, as in the JSON version of the results is:
{"columns":["path","path2"],"data":[],"stats":{"contains_updates":false,"nodes_created":0,"nodes_deleted":0,"properties_set":0,"relationships_created":0,"relationship_deleted":0,"labels_added":0,"labels_removed":0,"indexes_added":0,"indexes_removed":0,"constraints_added":0,"constraints_removed":0}}

Answer 1

This query of yours is using the same variable (p) for the Bob AND Susan nodes, which probably explains why it does not work as you expected (a single node cannot have 2 different values for the same property):
MATCH path=(p:Person {name: "Bob"})-[:reports_to*]->(o:Organization {code: "TopOfCompany"}),
path2=(p:Person {name: "Susan"})-[:reports_to*]->(o2:Organization {code: "TopOfCompany"})
RETURN path,path2;
You can either use different variables, or just get rid of the node variables entirely (since you don't use them anywhere) -- like this:
MATCH path=(:Person {name: "Bob"})-[:reports_to*]->(:Organization {code: "TopOfCompany"}),
path2=(:Person {name: "Susan"})-[:reports_to*]->(:Organization {code: "TopOfCompany"})
RETURN path,path2;

Answer 2

Optional Match, which could be considered the Cypher equivalent of outer join in SQL, can be used when working with path matching. The following query matches the two individual paths as well as the path that matches both people:
MATCH
path1=(p1:Person {name: "Bob"})-[:reports_to*]->(o1:Organization {code: "TopOfCompany"})
OPTIONAL MATCH
path2=(p2:Person {name: "Susan"})-[:reports_to*]->(o2:Organization {code: "TopOfCompany"})
OPTIONAL MATCH
path3=(p1)-[:reports_to*]->(o:Organization {code: "TopOfCompany"})<-[:reports_to*]-(p2)
RETURN path1, path2, path3;