Variable markersize in seaborn pointplot

I would like to vary the marker size in my seaborn pointplot according to a numpy array representing the sample size.

markersize_array = np.array([[10, 5, 6]])

Feeding markersize_array to the markersize keyword produces an error:

TypeError: only length-1 arrays can be converted to Python scalars

Here is some example code:

penguins = sns.load_dataset("penguins")

fig, ax = plt.subplots(1, 1, figsize=(2, 2))

markersize_array = np.array([[10, 5, 6]])

sns.pointplot(
    data=penguins,
    x="body_mass_g",
    y="island",
    ax=ax,
    estimator=np.mean,
    errorbar=('ci', 95),
    capsize=0,
    color="k",
    log_scale=False,
    linestyle="none",
    marker="s",
    markersize=np.array([[5]]),
    # markersize=markersize_array,  # <- I need this
)

Is there a way to make this happen?

Edit

Using seaborn objects also does not work:

import seaborn.objects as so

tips = sns.load_dataset("tips")

p1 = so.Plot(tips, "total_bill", "tip")
p1.add(so.Dot())
(
    so.Plot(tips, x="total_bill", y="day")
    .add(so.Dot(pointsize=np.array(10)), so.Agg())
    # .add(so.Dot(pointsize=np.array([[10, 10, 10, 10]])), so.Agg())
    .add(so.Range(), so.Est(errorbar=("se", 2)))
)

Maybe you can set the marker to "none" and use plt.scatter afterwards to draw the points at the right sizes.

import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns

penguins = sns.load_dataset("penguins")
x = "body_mass_g"
y = "island"

fig, ax = plt.subplots(figsize=(6, 4))

# Error bars
sns.pointplot(
    data=penguins,
    x=x,
    y=y,
    estimator="mean",
    errorbar=("ci", 95),
    capsize=0,
    color="k",
    ls="none",
    marker="s",  # "none" here
    ax=ax,
)

# Central points with variable sizes
means = penguins.groupby(y)[x].mean()
markersize_array = np.array([10, 5, 6])
ax.scatter(
    x=means.values,
    y=means.index,
    s=markersize_array * 40,
    color="green",
    marker="s",
    alpha=0.5,
)


fig.tight_layout()
fig.show()