In series cells, the weakest has the largest voltage mismatch error prone to over/under voltage damage during CV charge or low SOC.
In parallel cells the “strongest” battery has the largest current mismatch due to $$ΔI = Vc / ΔESR [mΩ]$$
Therefore all cells must be perfectly tested and matched before any strings series or shunt or both. Balancing extends the life of series cells. But for parallel cells, the strongest (highest C[kF] lowest ESR) is stressed more to become as weak as the shared one then the pair is compared to the rest of the series string for mismatch voltages for aging effect margin.
Shunting two strong cells of different chemistry voltage can lead to thermal problems due to circulating currents in formula above until balanced voltage and null circulating current. Then the current sharing depends on ESR mismatch. The “strength” of a battery can be computed by the set of ESR*C=T time constants (more than one). Lower is better for short term surge power, which is often rated by C rating * Ah (20h) where higher T is better for long term T’s for duration of Ah.
These are different capacitors inside the complex Battery equivalent circuit.(beyond the scope of this query.)
BTW your 1S pack is actually 1S12P. The bus bar is <<1 mΩ while each cell may be 50 mΩ +/- 50% range but ought to be matched <<1% in that range. You may test each with pulse load tests for ESR1 and steady avg I Amps to get ESR1 and C2 values, where C2=I dt/dV measured for typical duration expected (0.5h~tbd h) in a ESR1C1//ESR2C2 simplified 2 cap equivalent cct. This is based on the approx. asymptotic 2 slope V vs t curves.