Unable to edit words or phrases

The Japanese lingq sometimes fails to separate text into words properly. This is to be expected, discerning individual words in a Japanese text can be challenging even to human readers, and is certainly non-trivial for a computer. There is also the question of what part of the text should be considered a word, what would be a phrase and what is a mere part of a word - these things are not always clear-cut in Japanese.

The new phrase functionality has provided some crucial help in dealing with these issues. However, as I’ve just discovered, an important bit of functionality went missing: I am no longer able to edit a word.

I’ll give you an example. In the picture I’ve provided below, the following bit of text can be seen:

The correct way to split this (as clearly demonstrated by the last part of the picture which contains a Rikai-chan dictionary popup) would be
今 は ただ しょぼくれた 川縁

However, what lingq did was this:
今 は ただし ょぼ くれた 川縁

As I said, these things are bound to happen. In the old lingq, I’d simply select 「ただしょぼくれた」 and edit it down to just 「しょぼくれた」 It was tedious, but it worked. Since the new lingq enables phrases, I thought I’d just select 「しょぼくれた」 as a phrase. However, as I demonstrated in the picture, I was only able to select the phrase word by word, in other words I could select either 「ただしょぼくれた」 or 「ょぼくれた」, but not 「しょぼくれた」.

Situations like this seem pretty rare, but as you can see it does happen. I was only able to resolve this by editing the whole lesson, which is quite inconvenient and takes a lot of time. I would suggest the “edit word/phrase” function should be brought back.

I have a similar problem also in Chinese. When I import new lessons, it has trouble splitting or combining characters to make proper words, especially when it comes to names. For example: 摩拜单车, which is a brand of bike rental (English known as Mobike) is, not suprisingly, in google search because it is fairly new. That being said, the text I imported has both 摩拜 单车 and 磨 拜 单车 shown as is.

So similar to what GreenAirplane was asking, is there a way we can edit the spaces between characters in order for LingQ to effectively recognize words easier?