industry_increase_every = x means whenever a city reaches x*2^n citizens, a new end consumer spawns. It does not matter which end consumer that is. If you are lucky, a 10-factory-chain spawns at once. If you are unlucky, it's just a wind turbine doing nothing. Both take up one 'industry-increase-slot'.
I wonder if it would be feasable to add a parameter to end consumers that allows them to take up fractions of an industry-increase-slot, such that two or four smaller chains could spawn at the same time.
Basically, when a city reaches x*2^n citizens, the whole "build an industry" thingy fires. Nothing changes about that process, except once it's completed, it looks at the end consumer it just buildt. If that end consumer only accounts for eg. half an industry-increase-slot, the same "build an industry" routine starts over, except the pool of potential end consumers is limited to those who account for only half a slot or less, etc.
This would allow for smaller end consumers to be more viable in the game, rather than a disappointment. While a single Mom-and-pop corner store that needs a truckload of goods every other month is just incomparable to a large car dealership with a huge steel- and plastic industry behind it, if you can get several of them it's not so bad.
To keep it simple, powers of 2 should be enough - so half, quarter and an eight; maybe 2 could be implemented by storing a token that reminds to skips the next spawn round.
Note that this is completely independent of how industry_increase_every actually works. I know there were some talks/complains once about it being linked to population of individual cities rather than global population, but any changes in that regard would not interfere with the proposed concept.