I think I've found the answer to this question. Please correct me if I am wrong as I'm new to this field
As mentioned in the question, time to periapsis might be the key. From the Wikipedia description of Mean Anomaly, I found this equation
$$ M = n.(t-\tau) $$
$ M $ is Mean Anomaly, $ n $ is Mean angular motion, $ \tau $ is the time at which the body is at the periapsis. So $ (t-\tau) $ is the time to periapsis.
Mean anomaly(degrees) & Mean angular motion(revs/day) is given in the TLE. I just need to convert them into same units.
$$ (t-\tau) = \frac{M(deg).(\pi/180)}{n(rev/day).(2\pi/86400)} $$
This will give me time to periapsis in seconds for each satellite. So the distance between them, in terms of time, can be easily found by subtracting the two time to periapsis values.
To find the distance between two satellites in terms of space, I need to find the arc length of an ellipse. The equation provided by the answer is
$$ \int_0^{\theta_1} \sqrt{a^2.sin^2(\theta) + b^2.cos^2(\theta)} d\theta $$
$ a $ is semi-major axis, $ b $ is semi-minor axis & $ \theta_1 $ is the angle of the arc, which can be easily found using Mean angular motion.
I used WolfRam Alpha to compute the above equation & found the answer to the distance between 2 satellites in terms of space(km). I cross verified it using the general distance-time formula
$$ \frac{Orbit \ Circumference}{Orbit \ Period} = \frac{Arc \ Length}{Time \ to \ traverse \ Arc \ length}$$
Orbit Circumference can be found from semi-major, semi-minor axis & eccentricity. I used Google calculator. Orbit period can be found using Mean angular motion, Time to traverse Arc length is difference between time to periapsis for 2 Satellites, so the Distance between 2 Satellites in terms of space(km) is the only unknown.