Electric power is not zero emission - it's remote emission unless you have one of the vanishingly rare cases where the power comes entirely from a non-burning source like solar, wind, or hydro. Most electrically-powered transportaton relies on traditional burn-something power plants that are traditionally located in somebody else's back yard. On the plus side, they are in fact zero-LOCAL-emission so if there are more people near the transportation service than near the power plant it might work out somewhat better.

The push for Tier-4 comes from the air agencies as an alternative to electrification, and since the commuter locos are bought with tax money the cost doesn't really matter. And compare the cost of the custom locos to the cost of wires and (also custom) electric or dual-mode locos. Also, of course, the commuters are essentially local agencies with local politicians on the Boards so the usual federal preemption stuff either doesn't apply (not interstate commerce?) or is ignored for PR reasons. It's not really new - the Cal Amtrak F59phi's in the 1990s were essentially prototypes for Tier 1-2 engines and suffered the reliability consequences for quite a while.

EMD has no Tier-4 engines. The 710 series isn't capable of doing it without SCR aftertreatment which EMD and the railroads want to avoid. Cat engines are a reasonable alternative from a corporate standpoint since (with SCR, which commuter lines can handle more effectively than the Big RRs) they come from the same corporate family (EMD<Progressive Rail<Cat) - but to get the same HP you need 2 which then offers possible "genset" application. Does GE's Tier-4 EVO prototype do without SCR or did they give in and use it? MPI just uses whatever engine the customer specifies - they don't care as long as it fits, and for a price they will make anything fit.

As with rail transit gear, rail commute equipment is bought in small enough quantities so seldom that it's a custom job no matter what. The makers try to standardize major subassemblies but when purchases are 5-10 years apart and for only 5-10-15 units it'll never be mass production. And the operating requirements ARE different from freight so you can't just buy "off the shelf." That said - why does Metrolink need something suitable for speeds that won't be seen on the system at least within the life of the equipment if ever?