That image keeps popping up once and a while. It's complete and total bullshit and has been debunked many times.
Sex genes actually are established as soon as the pollen meets the calyx and the seed is developed (fem seed being the obvious proof). The actual "sexual expression" or "reproductive commitment" can be different from its sex genes, but it's not "strongly" influenced by the environment.
In general 50-50% is male/female. Getting 80/20 on 10 seeds doesn't change that, the ratio is only accurate over very large numbers.
The sexual differentiation of Cannabis sativa L.: A morphological and molecular study
V. M. Cristiana Moliterni, Luigi Cattivelli, P. Ranalli, Giuseppe Mandolino
2004
Summary
Cannabis sativa L. is a dioecious species with sexual dimorphism occurring in a late stage of plant development. Sex is determined by heteromorphic chromosomes (X and Y): male is the heterogametic sex (XY) and female is the homogametic one (XX). The sexual phenotype [trait...] of Cannabis often shows some flexibility leading to the differentiation of hermaphrodite flowers or bisexual inflorescences (monoecious phenotype). Sex is considered an important trait for hemp genetic improvement; therefore, the study of the mechanism of sexual differentiation is of paramount interest in hemp research. A morphological and molecular study of Cannabis sativa sexual differentiation has been carried out in the Italian dioecious cultivar Fibranova.
Microscopic analysis of male and female apices revealed that their reproductive commitment may occur as soon as the leaves of the fourth node emerge; the genetic expression of male and female apices at this stage has been compared by cDNA-AFLP. A rapid method for the early sex discrimination has been developed, based on the PCR amplification of a male-specific SCAR marker directly from a tissue fragment.
Five of the several cDNA-AFLP polymorphic fragments identified have been confirmed to be differentially expressed in male and female apices at the fourth node. Cloning and sequencing revealed that they belong to nine different mRNAs that were all induced in the female apices at this stage. Four out of them showed a high degree of similarity with known sequences: a putative permease, a SMT3-like protein, a putative kinesin and a RAC-GTP binding protein.
http://link.springer.com/article/10.1007/s10681-004-4758-7
