@bontwoody  - replying to your question on the article - its easier to reply in the forum.

 28mm plastic pipe inserts have a bore of about 21mm - so if you have lots of them in the path it matters. but  if you use engineeringtoolbox.com, you  work out the head loss for that specific length at that flow rate and bore size. then add up the head losses over the whole run. provided the section with small bore size and a higher head loss per m is very short, it contributes nothing significant to the the overall head loss  and thus achievable flow rate. whereas an undersized diverter valve or mag clean can burn 1m or more of head.  If you've only 5kw to transfer your target is lower so easier to achieve. be careful with the 15mm plastic - do the maths on throughput per radiator.

in my case after 1 week of running, I'm maxing at 30l/min. need to do some more testing to prove what section needs to be to be redone to push it higher. Steady state for a 21C target at 7C outside is  15l/min, 4.5kw output,  COP 3.0, LWT 38, so happy-ish for 1 week in. but I know there's a group of 4 rads fed by a 3m run of 22mm plastic that are getting lower flow temps than the rest (I've got monitors on everywhere), making it hard to balance the system. My overall DT is a little lower than I'd like (between 2 and 4, averaging 3) . and my strong suspicion is it's that 22m feed thats slowing the overall rate down, so next step will be to replumb that group when its warm enough to shut the system down.  I built a whole new primary side but there are bits of the secondary side that I know aren't up to spec.