Other people have already said a lot, but I'll fill in some more of the calculations. So, according to the poll, we only know that

P(Cis) = 0.28

and

P(Straight) = 0.19

Now, what we are looking after is P(Cis ∧ Straight). Since we don't know if cis people on this sub are more or less likely to be straight, there's no way to calculate this without making assumptions, but generally in statistics for a rough estimate we can assume statistical independence. In that case we get

P(Cis ∧ Straight) = P(Cis) * P(Straight) = 0.28 * 0.19 = 0.06

which would mean about 6% of people are cis and straight. That is probably underestimating it, because it is pretty likely that cis people are more likely to be straight, but from this data, there is no way to know.

Now, to what you calculated: instead of writing it in absolutes, you can rewrite it in probabilities:

P(Straight) / P(Cis)

In and of itself this gives us no information. But again, if we assume this time that all straight people are cis, which is a steeper assumption, we get the conditional probability:

P(Straight) / P(Cis) = P(Cis ∧ Straight) / P(Cis) = P(Straight | Cis) = 0.65

This gives us that assuming all straight people are cis, if you meet someone who is cis, there's a 65% chance they are also straight. Which is interesting, but not what we're looking for

wow, that turned out a bunch of nerd shit, what I actually meant to say was

:3