PowerBI.Dataflows vs PowerPlatform.Dataflows

After continued testing using identical query logic (after the navigation step) I can confirm the Power Platform Dataflow Connector is much slower than the Power BI Dataflow Connector, in some instances by 80%.

Below is my code (removed the navigation step ID's).

let
    Source = PowerPlatform.Dataflows(null),
    #"Bookings Detail" = *NAVIGATION STEPS*,
    #"Filtered Rows" = Table.SelectRows(#"Bookings Detail", each ([Closure Date Filter] = 0)),
    #"Removed Other Columns" = Table.SelectColumns(#"Filtered Rows",{"Centre Code", "Booking Date", "Booking Type", "Room Type", "Bookings"}),
    #"Replaced Value" = Table.ReplaceValue(#"Removed Other Columns",each [Room Type], each if [Room Type] = 1 then "Nursery" else if [Room Type] = 2 then "Toddler" else if [Room Type] = 3 then "Junior Kindy" else if [Room Type] = 4 then "Kindy" else if [Room Type] = 5 then "Pre-school" else if [Room Type] = 6 then "Before School Care" else if [Room Type] = 7 then "After School Care" else if [Room Type] = 8 then "Vacation Care" else if [Room Type] = 9 then "Before/After School Care" else null,Replacer.ReplaceValue,{"Room Type"}),
    #"Grouped Rows" = Table.Group(#"Replaced Value", {"Centre Code", "Booking Date", "Booking Type", "Room Type"}, {{"Bookings", each List.Sum([Bookings]), type nullable number}}),
    #"Added Conditional Column" = Table.AddColumn(#"Grouped Rows", "Room Type Sort", each if [Room Type] = "Toddler" then 2 else if [Room Type] = "Kindy" then 4 else if [Room Type] = "Pre-school" then 5 else if [Room Type] = "Nursery" then 1 else if [Room Type] = "Junior Kindy" then 3 else if [Room Type] = "After School Care" then 7 else if [Room Type] = "Before School Care" then 6 else if [Room Type] = "Vacation Care" then 9 else if [Room Type] = null then 10 else if [Room Type] = "Before/After School Care" then 8 else null, type number),
    #"Changed Type" = Table.TransformColumnTypes(#"Added Conditional Column",{{"Centre Code", type text}, {"Booking Date", type date}, {"Booking Type", type text}, {"Room Type", type text}, {"Bookings", Int64.Type}, {"Room Type Sort", Int64.Type}})
in
    #"Changed Type"

It appears the Power Platform Dataflow Connector loads the data by rows whereas the Power BI Dataflow Connector loads it by data and is much faster. It is worth mentioning I have a gigabit 1000/50 internet connection so I can download from Dataflows very fast.

It is very dissapointing that Microsoft have made the Power Platform Dataflow the default as clearly the performance is not as good.

It would be great to have an explanation as to why there is a big performance difference.