Question: will JWST catch up with L2 or will L2 catch up with JWST?
Answer: To a first approximation, both. And neither. The answer depends on the frame of reference you chose.
Case 1: In a rotating frame of reference (Sun and Earth stationary), the Sun, Earth and L2 are collinear so JWST follows a radial path from Earth towards L2.

Rotating frames of reference produce apparent violations of the laws of motion. Try playing ping pong on a merry-go-round for a demonstration.
In a non-rotating frame of reference, things appear different.
Case 2: In the Ecliptic Coordinate System (Sun centered, non-rotating), the L2 point is in a non-Keplerian, heliocentric “orbit”. It is outside Earth’s orbit, collinear with the Sun and Earth. Its heliocentric orbital period is one year. From this frame of reference, the transfer trajectory takes JWST 1,000,000 miles radially and 49,000,0000 miles circumferentially. Since JWST is in freefall, increasing its distance from both the Earth and the Sun , it must be slowing down. So, from this frame of reference, JWST is catching up to L2.

Case 3: In the Equatorial Coordinate System (Earth centered, non-rotating), L2 is in a 1 year, 1,000,000 mile orbit around Earth. L2 is traveling across the celestial sphere at 30$^\circ$/month. So in the (very approximately) 1 month from launch to L2 orbit, JWST appears to need to “catch up” 30$^\circ$ to J2. In this frame of reference, JWST is catching up to L2.
Fun facts:
L2 is at the meeting of Earth and Sun “spheres of gravitational influence”. So if you tried to model JWST trajectory with patched conic orbits (like Kerbal Space Program), it would fail.
The L2 point is in two different 1 year prograde “orbits” at the same time, around both Earth and the Sun.
Since L2s heliocentric “orbit” is 1% larger than the Earth’s, its orbital speed is 1% greater than the Earth’s and (roughly) 2% greater than an object in a Keplerian orbit at the same distance from the Sun.