To describe it, I'm not that guy. But to keep it short and sweet - your stacked tubes are the "secondary" towards the power transistors. The Winds are the Primary.
In light of transformers - you have "step up" - "step down" and Isolation - mean what they say- a way to also look at this is to say "impedance Network" "step up" - "step down" and Combination - again a way to look at taking one ohmic reactance result to another - it's done with pipes, ferrites and the winding methods used.
So if you think in relation to that - your length of pipe - folded over - is the winds your Base leads need - the Ferrite cores are a lot like the laminate and the winds thru the pipes - is where your power goes in.
Once the lengths of pipes and their chokes are done - the Secondary is done.
Primary? That is your tuning effort - the winds come from the combiner/splitter (input) to your Base's Pipe stack - the pipe stack is secondary, the winds are the primary to it - you can think of raising the impedance by raising the number of turns or winds you have to use to "meet" your expectations - it's science but also a labor of love - you may find yourself doing a lot of winds as trial and error - so best to use a template (Sigh) but also from that template comes the experience - and start out with more than enough winds - one or two extra - per "pipe and core stack". Lots of people think a tuning cap is the way to go, I wish I could say yes to that but what they miss is the direct coupling match tune they can obtain by working and doing the effort of hand tuning the stages and section of the amp. The efficiency of the mutual coupling without having the variable is satisfaction in itself.
Output side is the same way, only you work in reverse - you are taking the power transistors own output into a series of pipes (a transformer wind so to speak) and the wire winds within take up the energy from the pipes - and transfer it to the combiner - which is also taking in all the others. So you are doing a step up process - so you're taking the power developed from Collector To Emitter - thru the pipe that is also sending in DC voltage - and let the winds pick up the RF energy and since the winds are not directly connected to the DC voltage - they accept RF and take it's energy out of that stack of pipes and ferrites - leaving the DC behind for the power section of the transistors.
Now remember that you're dealing with proper impedance matching of the base to accept power and the output has to have it's level of impedance matching to develop and send that power out. So there will be a series of circulating currents of power running thru those output pipes - so it has to be correctly done so those currents are kept in phase relationship to each other - then that power is extracted off the pipes thru the winds. The output impedance the transistors work best in is slightly higher than the input impedance - so that is why the pipe lengths for input to output are different - its' a ratio. And why the pipes on the output side are or seem to be - twice as long as the input side - that's is for the very reason of propagating RF at a SPECIFIC range of frequencies - those pipes and ferrites and winds have to be a specific length to accept the energy developed without losing it in poor phase relationship (out of phase by a level of INCEPTION Start to End - sort of thing) so they appear as more like a tuning fork ringing at an octave of the other.
So in a way, you are making a series of or steps for - impedance - like a power transformer would for the input power - split, pared down into the sections - equally - then amplified - sent thru the output pipes and its' impedance - thru winds in those pipes, cores stack - back up to a step of impedance that allows the stage to be combined with others - summed into one single output of specific impedance.
A tuning cap is a quick and dirty way to fix a messy problem - but if all things are made equal - the effort is placed in similarity and duplication - then the problems go away or are easily solved because of the facts of one section goes bad does affect others but you arrive to the solution by seeing the results in your work and anything out of place? Just realign - put back in place and move on.
Another factor of "tuning cap blues" are just that - they cover up a messy stain or mis-match caused by poor wind and or tube lengths that were done for one type of transistor that is now no-longer made - that design is an orphan - best to start all over but for many that is not possible. So the tuning cap makes up for that orphaned design but you lose the power efficiency the amp design when properly done - can really do.
So much for "short and sweet"...
