A different approach to the first part: If $f(x)=0$ and $g(x)=0$, then also
$$0=f(x)+g(x)=(a-b)x+(b-a) $$
and this linear equation has exactly one root $x=1$ (where we use $a\ne b$!). So if they have a common root then it must be $1$. One verifies that $1$ is indeed a root of both.
Once we know one root of a quadratic, we can find the other by dividing out the corresponding linear factor. Or, recall that the product of the roots of $Ax^2+Bx+C$ (with $A\ne 0$) is $\frac CA$ (and the sum of the roots is $-\frac BA$). Knowing that one root is $=1$, the product of the roots is simply the other root. So the other root of $f$ is $\frac {c-a}{a-b}$ and the other root of $g$ is
$\frac{b-c}{b-a}$. So the desired result is that $\frac{a-c}{b-a}$ is between $\frac{b-a}{b-a}$ and $\frac{b-c}{b-a}$, or equivalently, $a-c$ is between $b-a$ and $b-c$. From $a<b$, we find $a-c<b-c$, hence what we still need is $b-a<a-c$, or equivalently
$$\tag1 b+c<2a.$$
Unfortunaltely,$(1)$ is not a consequence of the given inequalitites. In fact, if $b>a$, then $2a^2+b^2+ac<3ab+bc$ holds for all $c\gg0$, wheras $(1)$ certainly does not.
In concreto, we can use this to construct a counterexample:
Let $a=2$, $b=4$, $c=1$. Then $a<b$ and $2a^2+b^2+ac<3ab+bc$, but we have $$ \begin{align}f(x)=-2x^2+3x-1&\text{ with roots } \{\tfrac12,1\},\\ g(x)=2x^2-5x+3&\text{ with roots } \{1,\tfrac32\}.\end{align}$$