It wasn't a known issue (although it's known now!).
Turns out the value the 1.0 driver was writing was incorrect (it is
outside the valid range on .NET DateTime values). Instead of
253402300800000 it should have been writing 253402300799999. This was
fixed as part of the work done to allow the BsonDateTime class support
the *full* range of BSON DateTime values, which is much larger than
the range of .NET DateTime values (although with lower precision).
What was missed in making this fix was the impact on programs reading
existing DateTime.MaxValue values.
I've created a JIRA ticket for this issue:
https://jira.mongodb.org/browse/CSHARP-258
Probably can't be fixed until next week though.