Unlike 5e, D&D 3.5 (and, by extension, Pathfinder) doesn't include bounded accuracy in its design philosophy. As a result, high-level characters are able to accrue much higher bonuses to their various stats, including skill checks and saving throws. For example, most 5e characters will never see a skill check bonus higher than +11, while in Pathfinder it's possible to exceed that bonus even at level one, and high-level characters can easily attain skill bonuses of +30 or more. While the linked page doesn't seem to state what level of PC's it's intended for, it looks like it's built for a party of around level 8-10, perhaps slightly higher. Therefore, I'm going to evaluate these numbers in the context of what they'd mean to a level 10 Pathfinder party, and try to convert them into numbers that would have similar meaning to a level 10 5e party.
each PC needs to make a DC 24 Will Save
At this level, a DC 24 will save is still fairly difficult for many characters - a character specialized in the relevant stats (e.g. a cleric) would generally have a better than even chance of succeeding, but other characters would struggle. A typical fighter would likely fail the save around 75% of the time. Converting those probabilities into numbers for a different system is always somewhat inexact, but a DC 16 wisdom save would be a comparable 5e equivalent.
or suffer a −8 penalty to Perception checks, and a −4 penalty to all other actions
In 5e, it's atypical for circumstances to apply numerical bonuses or penalties. Instead, I'd suggest simplifying this mechanic to "disadvantage to all rolls on a failed save, no effect on a successful save".
Additionally, concentration checks must be made to cast spells successfully (DC 15 + spell level × 2)
In 5e, concentration checks are generally handled using constitution saves, but here it would likely be more appropriate to have them save using their primary spellcasting stat. In Pathfinder, the DC seen here is approximately a 50-60% chance of success, assuming the character is casting their highest-level spells. 5e doesn't generally scale DC's based on spell level, so it would be more appropriate to use a flat DC instead. Personally, I would use a DC of 16 here. Depending on how the numbers work out, it might have slightly higher odds of success than the Pathfinder version, but players no longer have the option to improve their odds by deliberately casting weaker spells, so I'd say it evens out. Plus, I'd like to keep the same DC across all saves against the effect, and this matches the DC used for the earlier wisdom save.
During combat, a DC 15 Perception check must be made to target an enemy, otherwise, the enemy appears to not be there when the attack is made, or if there is a friend within 5 feet of the intended target, that PC is targeted by the attack instead.
For a level 10 Pathfinder character, DC 15 is pretty trivial. When not impaired, most characters will auto-pass this check, or at least have very low odds of failure. However, since we're using disadvantage instead of a flat numerical penalty, we can't drive the DC too low. Assuming most of your characters are proficient in perception, DC 8 would be reasonable. If most are nonproficient, DC 6-7 might be more appropriate.
So, bringing it all together:
While in a visual patch of madness all light begins to strobe at irregular intervals. This strobing also affects darkvision. When the strobing starts, each PC needs to make a DC 16 Wisdom Save or suffer disadvantage on all d20 rolls. Additionally, casting a spell successfully requires a DC 16 save using the spellcaster's primary casting stat. During combat, a DC 8 Wisdom (Perception) check must be made to target an enemy, otherwise, the enemy appears to not be there when the attack is made, or if there is a friend within 5 feet of the intended target, that PC is targeted by the attack instead.