I think you are mostly right and phosphate buffer with pH 7 or 8 can be prepared using titration of KH2PO4 with KOH. Note, however, that if you take from the beginning, for example, 0.1M KH2PO4, the concentration of phosphate ions after addition of KOH will be lower than 0.1M. If you account for this dilution and dulute the buffer solution further to, for example, 0.05M phosphate concentration, everything should be fine.

Of course, I dilute the phosphate in for example 0.8l and after adjusting pH I add water up to 1l, so the dilution is not any issue.

You can do that, of course. I think it's just a matter of convenience and stability, once you have your mono- and dibasic 1M stock solutions, you simply mix the appropriate amounts of each to obtain a 1M solution at the pH you are interested in. You do not have to work with concentrated acids/bases, and do not have to worry about dilution.

Then it looks fine for me. You may have a very slight shift of pH after dilution, which you will need to correct with tiny amount of concentrated acid or alkali. These "iterations" is the only hindrance I can see. I am curious if somebody can correct us.

You need to prepare two solutions in the molarity of your choice (e.i. 0.1 mol/L).

1. 0.1 mol/L K2HPO4

and

2. 0.1 mol/L KH2PO4

mix them until the pH is at the level you want. This is the way to prepare any buffer, mixing the alkaline and the acetous components until the pH of choice is reached. If you are trying to adjust the pH using KOH of HPO4 you wouldn't have any constant molarity. If you know once how much of both components you need, you can go for recalculation of the salt components you used to get the amount of both for e.i. 1 L.