If the vessels (and the load distribution of whatever they're containing) are truly symmetric (or close enough), then yes, you can simply replace them with four equal concentrated loads at the points where they are supported.
If the vessels aren't symmetric but can be reasonably considered to have a pretty uniform stiffness (they don't have two really strong legs and two weak ones, for example), then you can use the vessels' centers of gravity to calculate the load distribution via this equation:
$$R = P\cdot\left(1-\dfrac{c_x}{L_x}\right)\cdot\left(1-\dfrac{c_y}{L_y}\right)$$
where $c_x$ and $c_y$ are the distance in the $x$ and $y$ axes, respectively, from the desired support to the center of gravity, and $L_x$ and $L_y$ are the spans between supports in the respective axes.
So, if the center of gravity is perfectly centered between all supports, then $\dfrac{c_x}{L_x}=\dfrac{c_y}{L_y}=\dfrac{1}{2}$ and $R = \dfrac{P}{4}$.
If, however, the center of gravity is at coordinate $\left(\frac{L}{3}, \frac{L}{3}\right)$ (so, close to the bottom-left support), then the load will be distributed $\frac{4}{9}$ at the bottom-left support, $\frac{2}{9}$ at the top-left and bottom-right supports, and only $\frac{1}{9}$ at the top-right support.
If the vessels are totally asymmetric, then you need to know the actual vessel geometry to calculate the load distributions.
This is a simple assumption which is perfectly correct for static analyses. If you need to deal with vibrations or other dynamic behaviors, then there may be fluctuations on how much load goes to each support.
