According to the most recent census there were 22,000 users of BSL as their "main" language and 65% of those could not understand English "well". So 14,000 people.
According to the most recent National Transport Survey, 41% of the population never travel by train. So 8,000 people.
We're spending £900,000 over three years on a national solution that at the absolute maximum will be of use to 0.012% of the population. Or put another way, we're spending £37 per year for every single person in the population who may use it.
In the same census data, there are 151,000 people in the UK who understand Polish "well" but do not understand English "well". With the same adjustment for people who never travel by train, 89,000 people.
So you'd get 11 times more benefit having every PIS in the country additionally in Polish even if it cost the same as BSL, which it wouldn't. Wonder why that doesn't happen?
It's gesture politics not proportionate to the benefit. If the real desire was to improve understanding of the PIS, there are 42 languages you would implement before BSL.