Sure. Including solution + explanation for comprehension's sake.

The problem is:

Get the row per [Number] and [Max Dat Reported] with closest earlier dateAdded relative to Max Date Reported. Logic for dateAdded vs Max Date Reported:

• target dateAdded must be equal to or earlier than Max Date Reported
  AND
• target dateAdded should have smallest difference with Max Date Reported

The solution is:

let
    Source = Original,
    RemoveAddedAfterMax = Table.SelectRows(
        Source, 
        each [dateAdded] <= DateTime.From([Max Date Reported])
    ),
    AddDiff = Table.AddColumn(
        RemoveAddedAfterMax, "Diff", 
        each DateTime.From([Max Date Reported]) - [dateAdded], 
        type duration
    ),
    SortDiff = Table.Sort(AddDiff, {{"Diff", Order.Ascending}}),
    GetSmallestDiffRow = Table.Group(
        SortDiff,
        {"contactId", "Max Date Reported"},
        {{"group", each Table.RemoveColumns(Table.FirstN(_, 1), "Diff"), Value.Type(Source)}}
    ),
    CombineSmallestDiffRows = Table.Combine(GetSmallestDiffRow[group])
in
    CombineSmallestDiffRows

And, as requested, here is an explanation on last two steps.

GetSmallestDiffRow:

The objective of the step is to find the row that meets our criteria by every [Number] and [Max Dat Reported] pair. In scenarios like this, Group By is the route to take.

First, consider the following simple "Group By" UI operation on the SortDiff step from the solution.

This produces the following

output (with preview of All Rows output per group row) 

and M code (check out here for good explanation on Table.Group).

Now, back to explaining solution code.

In both UI code and solution code, you can see we are taking SortDiff table and grouping by columns {Number, Max Date Reported}. The difference is the aggregation code in the third argument. Compare: 

In conclusion for this step, we are producing a grouped agg column "group" where each row is a single row table with the target "closest earlier dateAdded relative to Max Date Reported"

CombineSmallestDiffRows:

So, we have what we want from the previous step. The problem is that they are nested inside a table column. Rather than this column of separated target rows per [Number] and [Max Date Reported] pair, we want a single table that is the combination of all those single-row tables. Our objective for this step is to perform this transformation (column of tables -> single table combination of tables in column).

Again, let's first consider a simple UI example.

Take

Table1

and

Table2

Now, let's Append these through the UI.

Here is resulting output

and resulting code.

So, we append tables together with Table.Combine, which takes as its only argument a list of tables.

Now, back to the solution step.

As discussed just above, at a high level, the previous step added a column of tables. And when you reference a table column like "Table[Column]", the output is a list of the values of the column. In the case of our "group" column that we added, the values are tables (single target row). In other words, "GetSmallestDiffRow[group]" gives us a list tables, which is exactly the input for Table.Combine that we want in order to get our final output.