Here I got $X_2(t) = (t\cos t, t\sin t)$ where $(0 \le t \le 2\pi)$
Let $X_2(t) = (x, y)$ then $x^2 + y^2 = t^2 \;\;\forall t\in[0, 2\pi]$
now will divide $[0,2\pi]$ into $n-identical$ segments.
then we get $[0, \frac{2\pi}{n},2\frac{2\pi}{n},..k\frac{2\pi}{n}, (k+1)\frac{2\pi}{n},...n\frac{2\pi}{n} ]$
then we take $kth$ and $(k+1)th$ term, and measure the distance between two:
$d(X_2(k\frac{2\pi}{n}),X_2((k+1)\frac{2\pi}{n})) = \sqrt{[k\frac{2\pi}{n}\cos k\frac{2\pi}{n}-(k+1)\frac{2\pi}{n}\cos((k+1)\frac{2\pi}{n})]^2+[k\frac{2\pi}{n}\sin k\frac{2\pi}{n}-(k+1)\frac{2\pi}{n}\sin((k+1)\frac{2\pi}{n})]^2}=\sqrt{k^2(\frac{2\pi}{n})^2+(k+1)^2(\frac{2\pi}{n})^2-2k(k+1)(\frac{2\pi}{n})^2[\cos k\frac{2\pi}{n}\cos (k+1)\frac{2\pi}{n}+\sin k\frac{2\pi}{n}\sin (k+1)\frac{2\pi}{n}]} $
then if we let $n\rightarrow \infty$, we can estimate that $[\cos k\frac{2\pi}{n}\cos (k+1)\frac{2\pi}{n}+\sin k\frac{2\pi}{n}\sin (k+1)\frac{2\pi}{n}] \rightarrow 1$ thus $d(X_2(k\frac{2\pi}{n}),X_2((k+1)\frac{2\pi}{n})) \rightarrow \frac{2\pi}{n}$ thus, total length would be $\frac{2\pi}{n}\cdot n = 2\pi$
I had developed my own conjecture as denoted above, and wants to be verified and get some correction for illogical reasoning.
Any harsh ones are welcomed.
The arc length is equal to
$$\int_0^{2\pi}\sqrt{\left(\frac{dx}{dt}\right)^2+\left(\frac{dy}{dt}\right)^2}dt$$
We have $\displaystyle \frac{dx}{dt}=\cos t-t\sin t$ and $\displaystyle \frac{dy}{dt}=\sin t+t\cos t$
$$\left(\frac{dx}{dt}\right)^2+\left(\frac{dy}{dt}\right)^2=(\cos t-t\sin t)^2+(\sin t+t\cos t)^2=1+t^2$$
Let $t=\tan\theta$. The arc length is
\begin{align} \int_0^{2\pi}\sqrt{1+t^2}dt&=\int_{\theta=0}^{\tan\theta=2\pi}\sec\theta d\tan\theta\\ &=\Big[\sec\theta\tan\theta\Big]_{\theta=0}^{\tan\theta=2\pi}-\int_{\theta=0}^{\tan\theta=2\pi}\sec\theta\tan^2\theta d\theta\\ &=2\pi\sqrt{4\pi^2+1 }-\int_{\theta=0}^{\tan\theta=2\pi}(\sec^3\theta-\sec\theta) d\theta\\ &=2\pi\sqrt{4\pi^2+1 }-\int_{\theta=0}^{\tan\theta=2\pi}\sec\theta d\tan\theta+\int_{\theta=0}^{\tan\theta=2\pi}\sec\theta d\theta\\ &=\pi\sqrt{4\pi^2+1 }+\frac{1}{2}\int_{\theta=0}^{\tan\theta=2\pi}\sec\theta d\theta\\ &=\pi\sqrt{4\pi^2+1 }+\frac{1}{2}\Big[\ln(\sec\theta+\tan\theta)\Big]_{\theta=0}^{\tan\theta=2\pi}\\ &=\pi\sqrt{4\pi^2+1 }+\frac{1}{2}\ln(\sqrt{4\pi^2+1}+2\pi) \end{align}
The estimated arc length is not correct.
The estimate of OP is incorrect at one point
It is correct that
$d(X_2(k\frac{2\pi}{n}),X_2((k+1)\frac{2\pi}{n})) =\sqrt{k^2(\frac{2\pi}{n})^2+(k+1)^2(\frac{2\pi}{n})^2-2k(k+1)(\frac{2\pi}{n})^2[\cos k\frac{2\pi}{n}\cos (k+1)\frac{2\pi}{n}+\sin k\frac{2\pi}{n}\sin (k+1)\frac{2\pi}{n}]} $
Note that
$$\cos k\frac{2\pi}{n}\cos (k+1)\frac{2\pi}{n}+\sin k\frac{2\pi}{n}\sin (k+1)\frac{2\pi}{n}=\cos[(k+1)-k]\left(\frac{2\pi}{n}\right)=\cos\frac{2\pi}{n}$$
So,
$$d(X_2(k\frac{2\pi}{n}),X_2((k+1)\frac{2\pi}{n})) =\frac{2\pi}{n}\sqrt{k^2+(k+1)^2-2k(k+1)\cos\frac{2\pi}{n}} $$
The arc length is approximately $\displaystyle \sum_{k=0}^n\frac{2\pi}{n}\sqrt{k^2+(k+1)^2-2k(k+1)\cos\frac{2\pi}{n}}$.
We cannot replace $\sqrt{k^2+(k+1)^2-2k(k+1)\cos\frac{2\pi}{n}}$ by $\sqrt{k^2+(k+1)^2-2k(k+1)}$ as there are many terms. Although $\displaystyle \cos\frac{2\pi}{n}$ is close to $1$, the error cumulatives when there are many terms.