Sounds like you have done a lot of research, in general the performance of a heat pump takes a very long 2nd place compared to system design, so should be of secondary importance.

What you have described of your experience with installers in France is the same as a lot of installers here.

As you have highlighted from your research, if they want to install a buffer, run away, they want a simple swop out for your boiler with no issues that requires zero system design.

Your 16mm pipework is likely to be an issue, but until you find an installer that will actually do a decent review of your heating system you won't know.

If it were me I would keep talking to different installers until you find one that actually knows what they are doing.

Good luck

When I first read this I thought 'no way thats going to work', but then I thought about it more carefully and I think it might.  However Im not an expert, just someone with a degree in physics and a background in engineering who has spent 2 years plus studying this stuff.  So I know almost nothing, but can (hopefully) think.  Sometimes thats an advantage!

You are definitely right not to trust the installers particularly when they give you a scattergun explanation, which means that they dont actually know.

Lest start with the monotube, radiators wired in series.  Im working from first principles here, someone else may have more knowledge.

In essence what you have is 2 3kW distributed emitters one for each floor.  Im working on 6kW total loss and the assumption that the floor demands are about the same. The 4 way balancing valve presumably acts as a partial bypass but it feels like it would be a nightmare to get this balanced particularly if you also add in TRVs.  So to first order I think they should be treated as a whole, much like UFH is.

The rads will be at different temperatures, which complicates the calculations.  Presumably French plumbers are used to that, but it makes a much greater difference at a FT of say 45 than it does at an FT of say 70.  For example suppose there are 5 rads in series each with a DT of 1C making 5C in total, and you have an FT of 45.  A radiator at the far end will emit 13% more energy than a same sized radiator at position 1 in the series.  Not too bad actually (probably no more than the uncertainty in the heat demands of the rooms) but worth calculating.  This will also have a subtle effect on the WC curve (more so if you operate at a very low FT like 35).  Of course this is less important if the building is in practice open plan, ie doors between major rooms are generally left open.

So lets go back to the view that each floor is one 3kW emitter.  15mm is good for 2.75kW at DT5

HEATGEEK.COM "https://www.heatgeek.com/does-my-pipework-need-upgrading-for-a-heat-pump-with-cheat-sheet/."

Heat Pump Piping: Do You Really Need to Upgrade Your Pipes?

Get expert guidance on heat pump piping sizing! Our comprehensive guide covers everything you need to know about heat pump piping sizes.

  So with eg 5 rads on any one floor each would operate at DT1, you would have a total DT of 5, and your 16mm pipework would suffice provided you can take 22mm pipework to the point the two floors divide.  That is beginning to sound like the basis of a plan!

Now lets look at the 'reasons' for a 4 port buffer tank:

• "We don’t know how your pipe network will behave." - well you should if you did some simple maths like that I did above.  Perhaps you could even calculate the index circuit?  I would guess its unlikely to be a problem, the floor area isnt so large so pipe runs are presumably fairly modest.
• "It helps during defrost cycles." - so does a volumiser (2 port buffer), without any of the downsides
• "It ensures the system volume is sufficient for the heat pump." - same as above
• "It prevents the heat pump from short cycling." - same as above
• "It improves performance by maintaining a ΔT of 5K." If the heat pump wants to maintain a DT of 5 it will likely do so by modulating the water pump.  If it doesnt want to it need not and obviously the pump designers decided it wasnt necessary

So I dont think there is any convincing reason for a buffer tank, subject to confirming that the water pump has sufficient head (which is likely given the size of the property).

As to choice of manufacturer, @gary is right, system design matters more.  However:

@johnmo has a Panasonic (I think) and regularly raves about it.  They dont seem to be that common in the UK, I suspect because they dont have a well established dealer network.

Mitsubishi are one of the most popular brands and have a good reputation.

Vaillant also have a good reputation and have a particularly easy to use and flexible UI (at least to the extent I can tell from reading the various instruction manuals and my personal experience of Vaillant which is what I have).  However there has recently been some bad press about the 7kW model, excess compressor noise during defrost and about capacity in this thread on openergymonitor.

I cant really comment on Viessmann, but they seem to have a good reputation albeit they are expensive at least in the UK

TBH I doubt any of these are a bad choice, its the system design that matters.

Split vs mono is debatable.  The advantage of a split is you dont have water outside (unless its a 'hydrosplit' -  where you do) and the outdoor unit is usually smaller, but the plumber needs f-gas and there is some internal noise.  Our domestic plumbers dont generally have that and the commercial boys, who do, are only interested in A2A.  So splits are rare here.

Finally, as you are in France, you perhaps dont have to sweat quite so much as we do about efficiency, because your electric prices are so low.  

Hope that helps