A lot of FPS games still use a lot of that code, or later versions.

Using bitwise and to implement modulo for a power of two is common as dirt in C; I'd object to it being considered a readability issue. Arguments about specific compilers in specific modes optimizing equivalent code to it using the % operator are beside the point. Often the whole point of using a power of two range is so that you can use a bitwise and at some point, and if it ultimately ends up in a library, it's likely that the argument to the and won't be constant (defeating the optimization). There are other considerations; at some point, the data type being wrapped around with the modulo may end up in a bitfield someplace (more likely in networking or tight storage situations). Being explicit about how many bits it uses doesn't hurt in that case.

Using &, rather than % is important, if someone used signed types for index for example, then % would've generate division, while & would still keep masking it - the example uses unsigned.

Also even if it was unsigned, but % was used then people might have the idea that any value could be used, while if & was used - it's more clear (at least to me) that power of 2 - 1 mask is needed - it just brings the right message.

The only case where % would've been used in similar fashion is in a hash-table, where the number of elements in the current bucket is prime number.

It is not very clear from the writing but he analysed the QuakeWorld source. QuakeWorld is the internet optimised version. So this is for server and client having latency between them.

  "I happily dove into Quake World source code."