getting the cost based on category

Your source data is in rather unfortunate shape. Here are some transforms to make it usable:

Costs:

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WclTSUVLADpwDQnFLAoGhkbFSrA5eI/xSS8rzi7LxGmNkZAQ2xol8lxgZGSrFxgIA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Project = _t, #"                     Category " = _t, #"                         Cost" = _t]),
    TCN = Table.TransformColumnNames(Source,each Text.Trim(_)),
    #"Trimmed Text" = Table.TransformColumns(TCN,{{"Project", Text.Trim, type text},{"Category", Text.Trim, type text},{"Cost", Text.Trim, type text}}),
    #"Changed Type" = Table.TransformColumnTypes(#"Trimmed Text",{{"Cost", Currency.Type}})
in
    #"Changed Type"

CPU:

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WUnBU0lFSIAAM9EwJqQIpidWJVnIibJyBnhFh0yzApjkTY5oxYdPMlWJjAQ==", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Project = _t, #"                 Region1_CPU Value" = _t, #"                       Region2_CPU Value" = _t]),
        TCN = Table.TransformColumnNames(Source,each Text.Trim(_)),
    #"Unpivoted Other Columns" = Table.UnpivotOtherColumns(TCN, {"Project"}, "Region", "Value"),
    #"Trimmed Text" = Table.TransformColumns(#"Unpivoted Other Columns",{{"Project", Text.Trim, type text},{"Value", Text.Trim, type text}}),
    #"Split Column by Delimiter" = Table.SplitColumn(#"Trimmed Text", "Region", Splitter.SplitTextByDelimiter("_", QuoteStyle.Csv), {"Region", "Category"}),
    #"Changed Type" = Table.TransformColumnTypes(#"Split Column by Delimiter",{{"Value", type number}})
in
    #"Changed Type"

Network:

let
    Source = Table.FromRows(Json.Document(Binary.Decompress(Binary.FromText("i45WUnBU0lFSIAAM9IwJqTLQM1eK1YlWciJsnIGeEWHTLMCmORNjGnFuiwUA", BinaryEncoding.Base64), Compression.Deflate)), let _t = ((type nullable text) meta [Serialized.Text = true]) in type table [Project = _t, #"                 Region1_Network Value" = _t, #"                      Region2_Network Value" = _t]),
        TCN = Table.TransformColumnNames(Source,each Text.Trim(_)),
    #"Unpivoted Other Columns" = Table.UnpivotOtherColumns(TCN, {"Project"}, "Region", "Value"),
    #"Trimmed Text" = Table.TransformColumns(#"Unpivoted Other Columns",{{"Project", Text.Trim, type text},{"Value", Text.Trim, type text}}),
    #"Split Column by Delimiter" = Table.SplitColumn(#"Trimmed Text", "Region", Splitter.SplitTextByDelimiter("_", QuoteStyle.Csv), {"Region", "Category"}),
    #"Changed Type" = Table.TransformColumnTypes(#"Split Column by Delimiter",{{"Value", type number}})
in
    #"Changed Type"

This will then allow you to combine the fact tables

Values:

let
    Source = CPU & Network
in
    Source

Now you can load the data.  Normally you should have separate tables for projects, categories and regions to make it a better data model.

From there you can now calculate all the required values

As you can see the column totals are incorrect.  That can be fixed by adjusting the measure according to your needs (or by not showing the totals).

Total = 
var p = values(Projects[Project])
var r = VALUES(Regions[Region])
var c = VALUES(Categories[Category])
var a = crossjoin(p,r,c)
var b = ADDCOLUMNS(a,"co",var p=[Project] var c=[Category] return CALCULATE(sum(Costs[Cost]),Costs[Category]=c,Projects[Project]=p),
                     "va",var p=[Project] var r=[Region] var c=[Category] return CALCULATE(sum('Values'[Value]),Costs[Category]=c,Regions[Region]=r,Projects[Project]=p))
return sumx(b,[co]*[va])

see attached.