Actually, I quite often find that the first of these two is exactly what I want as it makes what I'm doing much more clear. Othertimes, I use the latter.

Horses for courses I guess, but I think that in most every case I can recall, I find that <code> if( $var == 1 and $var == 2 )} doSomething(); } <code>

is as clear if not clearer than nested ifs as it clearly demonstrates that doSomething is dependant upon both condition; removes the potential for inadvertant mis-inclusions; and clarifes through the reduced clutter of unnecessary punction and syntax.

as well as whether there are other things I want to do in the $var==1 case but only if $var2!=2.

That is a different case, and a) requires the separation of the tests; b) is clarified by that separation.

In the cited case, the artificial separation leads (me) to questions about what was really in the mind of the author when the constructed the code. Questions that serve to confuse rather than clarify; doubt rather than know.