TopBunk,
You will find that there is no mention of "Fuel" in the piece you quote.
You forget that even if the fuel flow is higher, covering the ground at 8nm or more per minute copared to 4nm per minute has an effect.
There is also the issue of engine overhaul costs which have to be factored. For many cases these days, time is more expensive than fuel.
Therefore I recomend that you try playing with the CI in your machine and see if the descent is always at Vmd regardless of which (fuel or time) is set to be more influential.
It is a compromise and as I said in simple no wind situations, ther most efficient operating practice is to descend as quickly as possible.
Have a look at your descent tables and see what descending at say 2000ft per minute burns from FL350 to SL and then check what the burn is at say 3000ft/min.
In the second case the total fuel used during the descent will be less. It will be even smaller at a ROD of 4000 ft per minute.
The time to descend will be less so for the descent, a steep high speed descent will use both less fuel and less engine time / airframe time / crew time etc etc etc. As you say, small saving per flight but over a large fleet over a year it can mount up.
As for the extra time in the cruise?
20 miles at cruise speed (480 TAS) will take about 2.5 minutes.
In that 2.5 minutes, the aircraft descending at Vmd will only be about 3 or 4000 ft lower but now behind due to the reducing TAS (further behind if it has encounted and increasing headwind / decreasing tailwind). So your descent at Vmd keeps you in the air longer - using fuel and using hours (engine / airframe / crew / schedule).
So Vmd will not always be the most efficient.
Don't forget that the largest airline operator in the UK (the one that makes a big profit) uses high frequency and low costs as major tools to make money. The frequency of flights is a big factor and consequently, time may be quite a significant cost issue.