
(all credit @ TZW_Guile)
Several years back, someone over in the ACA thread claimed Gotch (they of Rutubo)'s Gradius had more slowdown than the PCB, with comparison videos provided. (an unfortunately ill-tempered proffering, but the record is what it is)
I suspected they were in fact witnessing the heavy CPU load of 5x Laser fire, something easily noticed on the PCB. (and indeed on Salamander, Gradius II, and Gradius III's PCBs) When 5x Lasers are crossing the screen, Gradius will slow down significantly. If you hold the fire button, you will notice a distinctive ebb/flow pattern to the game speed.
Old videos I made back then, from the runs provided:
Spoiler
EDIT: The BigCore fight is interesting. PCB player's total onscreen time is ~17sec, with them leaning on 5x Laser throughout. ACA player's is significantly faster, ~13sec, with them not firing until the core is exposed. However, going strictly from core exposure to explosion, their times are closely comparable; around ~7sec. Both players are leaning on 5x Laser during this time.
> Stage 4. PCB player pulls way ahead early on, making landfall ~5sec in front, before gradually losing ground at the Inverted Volcano, eventually falling behind. At that point, the ACA player switches to Double and quickly gains a ~5sec lead. Tragically cut short when PCB player dies at the enemy rush. 3; Given both players lean equally hard on their Lasers, I'm guessing the swapping of place is down to the visibly differing firing patterns.
> Vertical Arrangement. ACA player above, PCB player below. Audio PCB-only, a handy marker of the player's staggered firing pattern. And who doesn't love Mikado commentary?
> Merged Footage. ACA player colour, PCB player monochrome; audio again PCB.
Please enable subtitles/captions for some inlined notes. Alternatively, please see below:
Spoiler
> ACA player arrives at the base slightly ahead, due to not enabling Laser until late in the Air Battle.
> Once at the base, with both players 5x Lasering, the gap stabilises for several seconds, both videos demonstrating the usual 5x Laser ebb/flow...
> ...until the pivotal moment, at 0m45s. Half of the PCB player's lasers strike the ground, terminating prematurely. This causes him to start firing in a staggered 2-on/3-off pattern for the next 45 seconds, massively reducing slowdown.
> Note the ACA player's scenery falling behind to the rhythm of his 5x lasers; plunging back while they're onscreen, stabilising when they're not.
> After the PCB player lets off the fire button, resetting his laser array to 5x, the gap stabilises, much as it did at the base arrival. After ~45sec of staggered Laser fire, the PCB player reaches the midboss ~8sec ahead.

However, it's always good to talk. Usually! What do you chaps think?