Satoshi even mentioned this in his whitepaper https://bitcoin.org/bitcoin.pdf
See page 8 where it states:
"q=0.3
z=0     P=1.0000000
z=5     P=0.1773523
z=10   P=0.0416605
"

that means those control 30% hashrate (q=0.3) can mine 5 consequent blocks (z=5) by 17% probability and even can mine 10 consecutive blocks (z=10) by 4%. so, your reported case of a z=7 consequent block can happen sometimes.

Your surprise is legit as it is widely assumed that no one would control more than 10% of the hashrate so 6 confirmation is sufficient. But, currently this is not the case and if you refer what satoshi suggested for such a case is :
"P < 0.001
q=0.10 z=5
q=0.15 z=8
q=0.20 z=11
q=0.25 z=15
q=0.30 z=24
"
which means if there is an entity with 30% of hashrate, a transaction is safe from double spending (its probability is less than 0.1% ) after 24 confirmations (z=24).

You are right to be surprised as "the mining pools" breaks the original bitcoin philosophy.

Bitcoin design already addresses this issue and contains necessary mechanisms to make bitcoin sustainable even after block rewards get stopped.
Block miner will receive transaction fees, and when it is economically not rewarding for the labor/cost out there, some miners naturally will stop mining. However, since Bitcoin network difficulty adjustment will take care of average block length to be 10 minutes, thus network difficulty will drop and mining cost will drop. Eventually, mining cost will be less than mining reward that can be collected from transactions fees. So, bitcoin protocol has the necessary adaptation capability to handle your concerned case. No worry