i just got finished with that problem, and mine is a little bit different than yours. for instance, on the polyfit lines, you don't need to make them quadratic, i just made mine linear. also, i didn't do parts a b and c because on the website (http://calclab.math.tamu.edu/~dmanuel/calclab/m151/Fall/2009c_matlab8.html) all it says is to use exponential fitting on the pairs of points. so it's a lot simpler than you're making it out to be. all in all, i got this:
'delete s275x05.txt; diary s275x05.txt
clear; clc; close all; echo on
year=[0:1];
pop=[728 906];
p=polyfit(year,log(pop),1)
plot(year,pop,'o','MarkerFaceColor','r')
hold on
x=[0:0.01:6];
y=exp(polyval(p,x));
plot(x,y)
hold off
year=[2:3];
pop=[1171 1608]
p=polyfit(year,log(pop),1)
plot(year,pop,'o','MarkerFaceColor','r')
hold on
x=[0:0.01:6];
y=exp(polyval(p,x));
plot(x,y)
hold off
year=[3:4];
pop=[1608 2517];
p=polyfit(year,log(pop),1)
plot(year,pop,'o','MarkerFaceColor','r')
hold on
x=[0:0.01:6];
y=exp(polyval(p,x));
plot(x,y)
hold off
year=[0:4];
pop=[728 906 1171 1608 2517];
p=polyfit(year,log(pop),1)
plot(year,pop,'o','MarkerFaceColor','r')
hold on
x=[0:0.01:6];
y=exp(polyval(p,x));
plot(x,y)
hold off
echo off; diary off'