I am not a mathematician, I'm a chemist who happened to also be a math enthusiast.
I was playing with error function the other day, and just out of curiosity compute these integrals using Desmos. $$ \int _{-\infty}^{\infty} e^{-(x^2+x+1)} dx \approx 0.837247915445$$ Then, I just add more terms to it $$ \int _{-\infty}^{\infty} e^{-(x^4+x^3+x^2+x+1)} dx \approx 0.710530779033$$ $$ \int _{-\infty}^{\infty} e^{-(x^6+x^5+x^4+x^3+x^2+x+1)} dx \approx 0.682543666557$$ $$ \int _{-\infty}^{\infty} e^{-(x^8+x^7+x^6+x^5+x^4+x^3+x^2+x+1)} dx \approx 0.671744903773$$And so on. At 20th term (order 40 polynomial), the result is $0.655405978856$
One of the thing that I realize is the difference between terms became smaller, and I think the limit is $\frac{\pi^{0.5}}{e} \approx 0.652049332173292$. But I have no idea why, or if this is even true. I hope someone can provide me with what the limit is.
The functions $$ f_n(x) = \exp(-(1+x+\cdots + x^{2n}) = \begin{cases}\exp\left( -\frac{x^{2n+1}-1}{x-1}\right) & \text{ if } x \ne 1 \\ \exp\bigl(-(2n+1)\bigr) & \text{ if } x = 1 \end{cases} $$
Therefore Beppo Levi's lemma aka “monotone convergence” can be applied on the intervals $(-\infty, -1)$, $[-1, 0]$, $[0, 1]$, and $[1, \infty)$ separately, and it follows that $$ \lim_{n \to \infty} \int_{-\infty}^\infty f_n(x) \,dx = \int_{-1}^1 e^{-1/(1-x)} \, dx \, . $$
According to Wolfram Alpha the value of that integral is $$ Ei\left(-\frac 12\right) + \frac{2}{\sqrt e} \approx 0.653288 $$ where $Ei(x)$ is the exponential integral.