The reason is exactly what Kuba mentioned above. Until last October or so, you only needed 1 vote to approve/reject suggested edits (and 2 majority on Stack Overflow). This was with the old review system, where there wasn't any extrinsic motivation for reviewing anything and people reviewed for the fun of it.
Then they introduced the new review system with the possibility of 6 new gold and silver badges (one for each review queue), and things kinda went crazy (see here for background). On other bigger sites where the queue is substantial, people simply kept clicking "Approve" just to race to the badge and were not paying attention to what was being edited. The motivation to review was no longer about improving the site, but to grind a badge. Even though you get the badge for rejecting edits (both are counted as 1 review), approving is the easier task, because it's 1 click, whereas rejecting takes 3 ("Reject" -> select reason -> OK).
Bumping up the number of votes required to approve/reject increased the chances of bad reviews not making it through (bad reviewers are now caught by audits). Most of these might not be applicable to our site and certainly not to you as a reviewer, but it's only a matter of time before we get to similar levels. This is also one of those network wide changes, so every site has the same rules (except SO, which needs 3 votes).
You have a third option in the review queue, which is "Improve". Most of the suggested edits these days are for minor changes or simply using your script change \[Alpha]
to α
. You can improve it to fix things that they haven't done, which then applies your edit on top of their edit and gets accepted immediately because you have edit privileges. On the "improve edit page", here's a checkbox that says "suggested edit helpful". You can check it if the edit was truly helpful and if not, uncheck it. Depending on whether its checked or not, the edit gets approved/rejected by the Community bot. However, please use this only to genuinely improve posts, not as a back route to approving/rejecting an edit quickly.