It's punching straight into solar-orbit from take-off (no intermediate Earth-orbiting before boosting onwards), which means its increasing velocity probably renders minor fluctuations in the upper-atmosphere far less significant, and a lot of the adjustment needed is just subtle aiming (especially while the boosters are still boosting).
But it'll need to keep full throttle (staging changes aside) to have a similar sort of fuel-efficiency as an idealised suicide-burn landing hopes to achieve, to even reach the needle that it hopes to thread (the first Earth-Moon encounter, passing in exactly the right way to emerge onwards in order to thread the
next needle, and the rest after that), without having to adjust too much using precious boost/manoevering fuel.
And it'll be the most massive during lift-off, from the classic "rocket-equation" of needing fuel to lift the fuel, fuel to lift the fuel to lift the fuel, etc. Making anything less than a ground-level storm (into which they wouldn't launch) a more insignificant impact upon the calculated momentum of trajectory... Or so I would think.
Can't speak for the one-second, but I could believe that they'd rather try to hit such a precise mark each day, or await an equivalent (slightly shifted) day on the day(s) after rather than submit to wasting any more of the initially lofted fuel thaj necessary by finding themselves with a subtly different intra-solar trajectory that needs to waste propellant that they could have saved for later (when other planned burns had reduced the craft-mass significantly, so even more 'sponteneity' could be achieved by the same amount of reserve).
(and an Android 'clone', more recently), where I built asparagus-like stacks that I'd launch at various times of day (local to the launch site; like midnight, noon, 6AM, 6PM, and other times) and seeing where the power to go interstellar (if launched 'outwards') would send me.
Even accounting for minor differences in the stage-separation process (not wanting to settle into an orbit, I could keep my pinkies off any yaw-controls), there could be surprising amounts of chaotic differences according to exactly when before/after 6PM (say) my launch would ultimately take the tip-top of my stack into a highly eliptical retrograde orbit. And that's without more than one gravitational influence (only actually passing through the lunar area of gravity counts, in the simplified physis model) or any actual winds (just straight and predictably procedual air-resistance).
But it was fun getting a rocket that worked consistently and then seeing where it got to for various launch-times. Which I did far too much! (When I wasn't designing my modular orbital structures, and trying to be consistent enough in reaching orbit to not need much more fuel to bring the components together. To at least top up the already launched fuel-tanks from the prior launches, then drop the latest elements back into the atmosphere again if I'd otherwise only be adding on tanks empty of all but fumes.)