This is a known protocol design bug: signatures were defined in such a way that the cost of validating a transaction with N signatures is proportional to N2 rather than N.
It has a known simple solution: limit the number of inputs of a transaction to some reasonable value, say 20 or 100, independent of the block size limit. That will keep the cost of verifying one transaction bounded
Most of that is Greek to me. How would a sharp Core supporter respond?
Oops, it is actually worse than N
2, apparently
Trying again: There is a complicated transaction that barely fits in a 1 MB block, that takes 30 seconds to validate. Someone built a transaction twice as big, filling a 2 MB block, that takes twenty times as long (600 seconds; not just twice as long (60 seconds) as one would expect.
Right now there is no limit to the complexity of one transaction, except that the whole transaction must fit in a 1 MB block. If the block size limit is increased, then one could issue the more complicated transactions like the one above, probably causing problems for miners.
But there is a known simple solution that (AFAIK) was used in BitcoinXT and/or BIP101.