I've noticed how in some experiments, scientists use say 40 women vs 30 men, or basically they split the people unevenly between the control and experimental. But doesn't that make a bad experiment, because I presume you want equal numbers of people in every single group (control and experimental)?
What about research you may have carried out?

It depends on how the study is designed, what the biological readout is and how it's affected by gender, if at all, and how easy it is to recruit study participants. Most people will do what is called a power calculation when designing a study to work out how many people they need for the study to give it the statistical power to pick up differences between study participants.