Can a strategy be profitable (forward testing) without being profitable across many currencies?  In other words if I have a good strategy for GBPUSD via FSB, but it fails on USDCHF and EURUSD, is that possible long term?

There's lots of reasons they might not follow the same rules.  If you find a strategy that works across multiples that's probably a positive for the flexibility of the strategy itself however lack of profitability across multiples (especially without reoptimizing the parameters) is probably more common than not.  For instance if one currency is trending and the other is sideways it's unlikely the same strategy is going to work.  Also any fixed price distances, eg tp/sl and any indicator offsets will likely not be compatible (some of that can be worked around.)  A variance in spread between currencies can certainly make a difference in profitability as well.  If you've got any time range based indicators they'll also be off if for instance one currency is more active in the asian session and another during london.