I had a new Audi a4 back in 2007 with a 2.0 TDI 170hp engine - they were renowned for having failing injectors - the piezo coils used to give up.
It twice stopped in the outside lane of a motorway with no warning, whilst my wife was driving. She managed to cruise to the hard shoulder on both occasions thankfully.
After the first instance I had one injector changed - ouch
The second time I had the remaining three changed, the total cost for the four was over £2k - luckily a while after I’d paid this bill, VW / Audi admitted a problem with them and refunded the cost - apparently the high Dpf burn off temperatures were killing the injectors, frying the coils.
As for modern garages and so called fitters - if the computer says nothing, they are generally stumped - things are far too complex, and randomly changing bits gets you nothing other than a very big bill.

This is why I use a local garage that I trust. They use knowledge as well as what the computer says.

Update for today.

Garage confirms that they have tried a new egr and it made no difference. No surprise there then. I tried to tell them that but they are the experts. Original egr back in car.
Their next thought is the high pressure pump. I have tried to explain that the fuel pressure builds up almost instantly and stays within limits. I also said that if the pump was failing there would be metal particles in the fuel system. It was agreed that this was correct.
I now want the car put back together so that I can collect it tomorrow. I am going to need it over Christmas and the MOT is due shortly.
I'll have to work something out with them when I get there.
Thing is, although the injectors failed the test and new one's fitted, I've been told the issue is still present.
Car has been there three and a half weeks and I'm no further forward.  Now at the point when I learn something new something old is lost out the other side !
Now retired so it doesn't matter anymore.
Freelander now gone.

Surely the pressure of the high pressure pump can be read in live data

It can and I have.
I expect to pick it up tomorrow, complete with the original issue.  Now at the point when I learn something new something old is lost out the other side !
Now retired so it doesn't matter anymore.
Freelander now gone.

New Injectors don't solve the problem.
You know common rail pressure comes up just fine.
You know that the ECU is firing the injectors
I think you know that when it does fire , there is no lumpy running and it ticks over just fine
You know that it isn't air flow ( I think you said it ran without the throttle body connected)
I'm back at the ECU isn't commanding a sufficient injector pulse (for whatever reason). I thought a mobile injector specialist would have diagnosis cables to attach in-situ to the car , to actually record what the injector pulses look like.
(That would also show what happened to the pulse relative to the battery earth terminal , and any resultant starter cranking changes)

Hi Andy
Thanks for that piece of information, I didn't know you could do that. That will be something to mention when I take it back in a couple of weeks.
Took it for a run out yesterday and as predicted the hard start is still there. Unfortunately it now has a new issue. When the engine is under load climbing a long hill I get engine vibration through the steering wheel and it can be felt at the accelerator pedal. Also the fuel consumption is still poor.

Garage now wants to change the high pressure fuel pump to see if that makes a difference.
I' will have to try and reason with them again and mention what you suggest.  Now at the point when I learn something new something old is lost out the other side !
Now retired so it doesn't matter anymore.
Freelander now gone.

still need to remember that the rail pressures need to be within a specific threshold before the ECU will instigate the injectors.
If the pressure is too low or too high it will switch the injectors off to safe guard the engine.
i would suggest rather than throwing parts at it, it needs someone with a scope to go through it in some depth.
what is the battery voltage whilst cranking ?
What is the voltage between the negative battery terminal and the bell housing whilst cranking ?
re-reading all the posts, has someone scoped the camshaft signal for definitive output and check it back at ECU,
cam sensor signal determines start of injection.
I have had vehicles with too much back EMF from a starter motor interfere with cam sensor signal.
Scoping the starter for excessive back EMF is always good, had a transit about a year ago, everyone thought i was loopy but after another garage had fitted pump and injectors etc it was the starter causing interference with cam sensor. Jaguar x-type sport gone
Jaguar x-type 2.2d sport one of my best cars ever sadly gone
Freelander 2 GS auto 2008 most unreliable ever ! gladly gone
Freelander 2 GS manual 2013 only time will tell !

Hi Oldgeezer
I don't know if the garage has scoped anything, I certainly haven't as I don't have one.
As for some of your other observations.
Battery voltage is fine. Battery only couple of months old.
Have additional earth straps fitted.
Starter motor fairly new, approximately three months.
I know there is some form of signal going to the injectors during cranking as I used a noid light to confirm.

Fault has got to be something that only happens when the engine has cooled down. Engine starts fine when warm.
Never any rough idle or rough running after either start up.

I am still convinced it's a fueling issue.
ECU thinks the engine is warm, when in fact it is cold. Therefore does not inject enough fuel for a cold start.
All the relevant sensors have been changed. Therefore the ECU must be reading the information incorrectly.
If the ECU is not reporting any fault codes, it must be at fault.

That is my logic anyway.

Still open to suggestions though.  Now at the point when I learn something new something old is lost out the other side !
Now retired so it doesn't matter anymore.
Freelander now gone.

I am going to put my head above the trench

Could it be a faulty Coolant temp sensor ?

Have you done a live value check on it?

Consider head just shot off

Coolant temp sensor was replaced recently. Not because the old one was faulty but just like you I thought replace it anyway as it sounded like a good idea at the time.
Sadly, it made no difference.

As always I appreciate the input though.  Now at the point when I learn something new something old is lost out the other side !
Now retired so it doesn't matter anymore.
Freelander now gone.

So you postulate that the ECU is viewing the engine as warm when it actually its cold and hence poor cold starting. Would not the displayed coolant temp be wrong also? Just guessing.

Would it help to put the temp sensor in water of known temp (with a thermometer) and see what temp is registered via the OBD2 socket as the presumably analogue signal from the temp sensor is converted to digital somewhere ? Jules

Just a thought - the FL2 has an oil temperature sensor (LR001368) could this be part of the problem?

Please ignore if it's already been discussed - feeling lazy and couldn't be bothered to re-read all the posts - Gasman's been struggling for a while now and it's got be be getting annoying. Tangiers Orange - gone, missing her
Replaced by Ewok what a mistake - now a happy Disco Sport owner

There ought to be a prize for the best guess (or even the right answer)

I, for one, am very curious to know how this ends.

Good Luck Gasman I hope 2022 brings you a fully reliable FL2. FL2 XS SD4 Auto 2010 2012-2017 (21k - 91k miles) (MY2011)
FL2 Metropolis SD4 Auto 2014 2017- (16k - 77k+ miles) (MY2015)
Metro in its 11th Year of (Extended) LR Warranty / Full LR Service History
(Expensive, but Trouble/Worry free - hopefully?)

You need to check all the temperature sensors are giving the ecu a roughly correct reading this needs a good look at the live data its not always just enough to put new sensors in as it can be a wiring issue
It may be that a wire has rubbed through just enough to give an occasional bad reading
If the ecu according to live data is seeing every sensor as roughly the correct temperature and its still slow to start then you can rule temp sensors out but remember it uses quite a lot of sensors in its logic including the ambient temperature