Usually just slight flicking of the tube will be more than adequate for sample mixing. Unless your sample is highly viscous. Vortexing will definitely mix your sample, while at the same time introducing a lot of air/molecular oxygen and coating the sides of the tubes with sample. The spin down will help recover most of the sample at the bottom of the tube. Vortexing multiple times will not further mix a homogenous.

As Matthew says, you don't need much for mixing the samples. For example we would often just drag the 200 uL tubes across the rack to "shake up" but without actually shaking it.

Spin down is important because when mixing (whichever way you do it, vortexing or flicking or shaking) you often get small droplets on the walls/caps of the tubes and if you don't return these droplets to the bottom of the tube, you are risking sample loss. Even if you don't see the droplets they may still be there - a few small droplets in the tube lid can go unnoticed, and then there's the question of how much is stuck coating the tube wall. Especially when you have expensive stuff or limited amount of sample, you don't want to risk losing it!

Although not necessary, it is safe to mix cDNA by using a vortex briefly - once or twice makes no difference. Simply flicking the tube several times mixes cDNA (or any soluble molecules) quite well at normal working concentrations. If concentration is very high and the solution is viscous, gently pipetting up and down works well (ideally, you do not want the solution to be viscous - it is more likely to make significant errors in pipetting very small volumes).

If you are working with large size DNA (very large cDNA, plasmids, cosmids, or genomic DNA), try to avoid using a vortex. The larger the size, the greater the possibility of damaging the DNA by shearing. 

Dear Matthew Dunn, Marusa Bratus, Tausif Alam,

Thank you for your all great replies!

I learned a lot thanks to your help.