That is a very kind offer. I am not sure however how to post the proposals without broadcasting my personal details etc on the hub. HG is password protected too.

The house has 5 bedrooms and floor area of around 230 sq m, cavity walls insulated, double glazing and 200-250mm of roof insulation.

OK.  For comparison my house is 200sq m, 250mm loft insulation, a mix of  solid wall with IWI, uninsulated solid wall and lightweight block, double glazing, uninsulated floors.  7kW.  It is fairly close to cubic  which obviously helps.

Based on this and the scattergram below from heatpumpmonitor I think its quite likely yours is actually, <10kW, but whether its low enough for the Vaillant 7kW machine (or anyone else's ~8kW)) is clearly rather doubtful, and the 10kW Vaillant is just the 12kW downrated so not much point in that.  Thus I can understand the choice.  Since Heat Geek are guaranteeing performance, if it turns out they have seriously overestimated loss and it manifests itself in poor performance, they are presumably on the hook to do something.  Perhaps another reason to choose Heat Geek.

That is a very kind offer. I am not sure however how to post the proposals without broadcasting my personal details etc on the hub. HG is password protected too.

The house has 5 bedrooms and floor area of around 230 sq m, cavity walls insulated, double glazing and 200-250mm of roof insulation.

I’ve just sent a more detailed email to the editor looking for recommended installers for ASHP system around Rossendale in Lancashire, but if anyone else out there has experience or recommendations they would be gratefully received. 
we have already approached Global Energy as they are Lancashire based and manufacture here but having read a few threads on this forum they seem to get mixed reviews.

10kW is a high loss, is your house big/poorly insulated?  

Rad upgrades are an easy retrofit so if a few turn out to be too small it's not a big deal.  Age doesn't matter, size is what counts.

If hg are cheaper and genuinely guaranteeing cop of 3.9 and fitting vaillant as opposed to Grant I would probably go with them personally.  Better controls with a guarantee of performance.  Nothing wrong with grant although historically they have encouraged poor system design practice.  Guaranteed performance means that if rads are undersized they would have to up size them (depending on the precise terms!). 

I am beginning to wonder if HG play a numbers game, promising only the lowest likely number of rad upgrades but adding a bit of a margin to the price, then if it turns out they need more (which, in reality, you can only really tell by experiment) there is enough slack overall in the quotes (averaged over many instals) that they can afford to come back.  There would be some logic in this approach given the uncertainties involved!

Of course I don't have the details so this is based on limited info.  Has either given a detailed breakdown, if so I'm happy to take a look if you post them.

What flow temperature are they each designing for and what house loss are they estimating?  This will likely explain the difference.