Did you read the T's RFI documents? (I'll concede that I didn't, I don't even know if they made them public.) If you did and can point to where they may have put the thumb on the scale in favor of bi-levels, by all means do so (it'd be useful to know what to look for if they were to do such a thing again; they tend to get somewhat less effective at sandbagging when they're on their later attempts).
For convenience, the EMU RFI presentation slides. (Best source I can find in a two-minute search.)
More to the point,
no one who responded to the RFI bar Rotem has an existing FRA-compliant single-level EMU design. Alstom (though not their primary offering), CRRC, Hitachi, Rotem, and Stadler (also not their primary offering) at least mentioned single-levels, CRRC, Hitachi, and Rotem's proposed offerings were
only single-levels. None of them FRA-compliant. They'll all
happily build us one.
We're the ones who might wind up
very unhappy when, yet again, we're forced to be the guinea pig for a new design by an untested-in-the-US manufacturer (or, worse, Rotem, who's been tested and found basically incompetent).
No one (strawmen arguments don't count) has ever suggested that it's impossible to get a single-level FRA-compliant EMU. It's entirely doable, and the manufacturers would, indeed, provide viable ones in a heartbeat as you suggest. They'd also charge a ludicrous amount for them, because there's a metric shit-ton of work and risk that goes into modding a design for even the FRA's-newer, slightly-less-backwards regulations. Bombardier isn't having the fastest or easiest time with the MLV EMU, and that's just stuffing EMU guts into the existing FRA-compliant MLV, on a system they're very familiar with given that NJT's locomotive orders are near-enough sole-source BBD at this point.
The point I have been trying, repeatedly, to make is that the T is obliged (both out of good management practice and, more directly, by procurement rules) to make equipment decisions based on more than idealistic-perfectionistic vision of what the fastest-boarding EMU is. Something bi-level like the MLV or the KISS would presumably score lower on metrics involving speed-of-boarding, but if it's weighing that and say, them wasting their capacity to some degree on some routes versus them scoring a lot higher on risk and cost (because, remember, NJT and California are eating the first-adopter tax for a change), then that works out to be the more sensible purpose, even if it leaves the flat's better-boarding perfectionists pulling their hair out. That RFI's obviously a bit out of date now, and we'll see what's on offer at the next bite of the apple, but from the actual documents there were six options: two bilevels where someone else is eating the compliance-modification costs and the teething problems (California KISS and MLV EMU), three vaporware imports we'd have to eat those costs on (Alstom, CRRC, Hitachi), and one FRA-compliant single-level lemon (Rotem) that we shouldn't touch with a ten-foot pole. Unless someone else decides they want to be the beta tester for a Euro-import between now and then, the next RFI's going to turn up basically the same things. It's not about whether the manufacturers can or will offer a single-level, it's entirely about whether the sum total of factors it would take to get them works out favorably in comparison to getting the probably-much-easier-to-acquire bilevels. (If you ask me, "best fit at best cost" is a sea change in the right direction compared to the T's too-frequent habit of overcustomization.)