always 50/50... xx x xy = 2 xy and 2xx.
simple permutation... .5^x where x is the number of coin flips. so .5^4 * 100 = 6.25% chance all are male or female, if germinating 4. (.5 x .5 x .5 ...50%, 25%, 12.5%. 6.25%... etc... keep halving it for as long as you want)
anyone who says otherwise on the 50/50 thing is outing themselves as uneducated about biology/genetics in even the most basic understanding and simply pulling shit from their asses when the knowledge exists otherwise. the plant has 2 sex chomosomes. sex gametes are formed through meosis. this will split up the genetic material evenly between two new cells created for sexual reproduction. under normal conditions, all ovaries will have 1 x chromosome, since the parent has 2 x's. All pollen will contain 1x or 1y chormosome evenly split, because it came from diploid cells with xy paring.
pollen with 1 x or 1 y does not have any unique abilities that make it ore likely to find a pistil and knock up an ovary. so, it's seemingly random, although technically totally predictable with the right data collection and resolution. life is just dominos falling into the next domino...
so, those sources that said it may not be 50/50/... you should suspect anything they say in regard to real science... because they overtly make things up when that knowledge exists and has existed for centuries