patience.. keep it healthy. The rest is genetics and canopy management. By week 6-7 they should look the part but still need ripening. (if counting flower weeks from preflower, add 1 to compare apples to apples with photoperiod development). you don't actually know the point an autoflower flips to flower phase. you ahve to guesstimate based on development, which makes it inherently inaccurate.
The plant in the first picture is too lush (dark). Something is building up in the leaves - N or Mg are the candidates that cause this sort of symptom. Before the leaves take a downturn, i'd adjust your fertilizer formula once you rule out one of the possibilities.
genetics is the potential, but you can influence size of colas relative to genetic potential, too. E.g. too many colas per sq ft results in smaller buds. As long as you have 'enough' colas per sq ft, any addition beyond that merely spreads out the same mass of buds across more numerous colas -- smaller buds result. How deeply you get good buds in canopy is impacted too, as yield is spread out more. Light absorption = mass of flower not number of colas (assumes minimum number to reach max yield).
Shoot for ~3 colas per sq ft and you'll avoid more larf than someone overcrowding the space with colas.