suppose we have a code base with lots of legacy code using Crystal Reports. Now, the fairy godmother shows up offering the Fairy Reports ™ that will make it much cheaper and faster to make a new report given a spec than what development now costs us. But, this only applies to new reports, and we still need to maintain legacy reports as usual.
In this situation are there known arguments for why a switch to the new framework for new reports would be a bad idea? Or is this in fact precisely the strategy successfully undertaken by lots of companies out there where management has decided that Crystal Reports is worse than an existing alternative?
A somewhat related question is, just how much legacy maintenance do reports (Crystal or otherwise) usually need? Is more effort usually spent altering existing reports or creating new ones? How radical are the typical reports alterations?