Sounds like a classic case of "let's not make a future upgrade impossible". The material cost itself is only a small part of the total, so making the new bridges slightly higher is a rounding error in your budget. However, you're saving many millions if there were ever a full-line upgrade in the future, as you no longer need to do a full bridge replacement during that upgrade.

As long as there is even the vaguest plan of an upgrade at some point in the future it makes sense to adhere to the new standard, just in case. It's the no-regret option.

The costs would be high and the benefits negligible. Where more capacity is required, it's typically easier to lengthen existing trains, or run more trains, than to go adapting stations and building new unique double-decker trains that are only going to be compatible with a specific line.

And, as mentioned elsewhere, double-deckers have a big disadvantage in lengthening dwell times (due to less doors per passenger), which could result in slower services.