a Operations sequences seq5 and seqR applied to model. Pyramid graph shows how stack elongates/truncates over time (total signals on stack) under each sequence (ideal operation). Heat maps b and d show how stack pop limit is sensitive to washing procedure efficiency for seq5 and seqR respectively. Pop limit is denoted by overlaid numbers on heat map surface contours. Each heatmap square is a single simulation. Heat maps c and e show how pop limits vary at points W1 and W3 on (b) and (d), respectively, when strand concentrations λ and reaction wait times tw deviate from standard conditions 300 nM and 30 min. Dots on heatmaps indicate position of standard conditions. Note the log scales. f How pop limit is affected by pipetting noise under W1 washing. Pipetted concentration is drawn from uniform distribution 300 nM ± η. Average pop limit (lines) calculated from n = 25 independent stochastic simulations performed at each noise value, standard deviation shown as line shadows. g Same as f but using W3 efficient washing: overlapping lines seqR, seq10, and seq20 omit standard deviation shadows (drawn in Supplementary Note 10.4).