But I've read about miners which signaling on bit 1 has been declined after BIP91 locked-in, maybe you can see it on bitcoin core wallet or it just fud news?
That is completely incorrect. Signaling on bit 1 has significantly increased after BIP 91 locked in. You can clearly see the increase in signaling of bit 1 over the past day in this graph:
The yellow line is segwit signaling over the last 144 blocks. As you can see, it has sharply increased today as BIP 91 locked in.
I'm wondering too, what's going on. If they don't use bit 1 or BTC1 (v1.14.4) software, is it matter? As segwit2x should be activated in 2 weeks, right?
No. BIP 91 (aka the first half of segwit2x) requires that all blocks after activation must signal for segwit on bit 1. Segwit2x is not a different implementation of segwit. Rather it is a plan to activate segwit via BIP 91 (and BIP 91 activates segwit by requiring all blocks signal for segwit) and then to hard fork later.
I'm waiting for BIP148 which is on mooting right now, whether it still necessary or whatever it could be.
Right now, BIP 148 is theoretically pointless as miners should be enforcing BIP 91. If they do not, then BIP 148 will have an effect as then BIP 148's rules activate on August 1st. However I am told that most miners are running either btc1 (aka the segwit2x client) or segsignal (aka a BIP 91 patch on top of Core) so they will be enforcing BIP 91.
Encapsulated means another way, a secret path to activate segwit
BIP91 includes a hard forking, about 2Mb hard fork, right?
No, BIP 91 does not include a hard fork. That is segwit2x. Segwit2x contains BIP 91, but BIP 91 is not segwit2x.
Does hard fork even really necessary? Does bitcoin need 2Mb blocksize?
Calling it a 2 MB block size is actually completely wrong. The maximum block size would be 8 MB. It does represent a capacity increase, but it may be one that is unnecessary with all of the things that segwit brings us. In fact, such a hard fork could be harmful.