CAT tools - are they really helpful???

Last spring I bought one CAT tool (memoQ) and although I it tried many times already on different files, so far it has proven "helpful" in just ONE case. Everything else has been trouble!
Most probably this is because I am just too stupid, but I suspect some more "basic" reasons too.

For example:
I get a Word file for partial translation = meaning the file contains German, English and Japanese text, of which only the Japanese portions scattered throughout the document need to be translated.
1) I have to import the file.
2) Select all German/English text portions and copy them into the target segement.
3) Translate a few segments, want to take a break or continue working somewhere else.
4) Attempt BACKUP the "project".
=> Get error messages that tell me, I cannot do that. The reasons elude me.
*  In fall of last year someone took remote control of my computer to resolve similar problems with an Excel file, and commented: "Oh, that is a "known issue" = bug. I will try to make the developers fix this."
5) Attempt to "export" the active document, which still contains a few hundred untranslated segments, fails too. Error message:
"At least one of the documents you are trying to export contains errors than prevent memoQ from exporting the document.
Do you wish to resolve these errors now?" Yes ->
6) Results, naturally, a very long list of items/segments, where the "error" is:
"Too few memoQ {tags} in target segment"
This is a matter of course too, since I have yet done anything with those segments!

The customer support tells me:
"It is not possible to simply ignore all the tag errors in memoQ, and export a monolingual document in it's original format, as the missing tags would make that impossible."
7) "Export bilingual" sort of "works" = it gives me a 2-column, intricately interlaced and completely garbled rtf file in which I would have to search and than manually copy each and every already translated segment to the orginal Word file. (Since memoQ cannot handle ODF files, I am also stuck with Word.)

Well, THAT does not make ANY sense. In that case I would be a lot better off using the original file from the beginning and forget about memoQ altogether.
Actually, that was the idea when I tried to "export" the document.

So, with the mentioned ONE SINGLE exception I ALWAYS get the some (or all) of the following problems
* Software cannot import original file
* Software cannot export imported file
* Software cannot "backup" the project
* In order to work on the designated text, I have to work on ALL OTHER text too.* The work put into a file only half finished apparently cannot be recovered be recovered in any sensible/practical way.
* I always MANUALLY have to tell memoQ, that it should ignore various things I would NEVER have to care about working with an ordinary (Word/text) file. There I would for example just a date/number and that's it. In memoQ I am confronted with the need to check "number format", verify numbers, insert tags etc. etc.
What in a word processor is for example TWO keystrokes (for example "10") may well need to required 10 or more "clicks" to get the same thing done!
That is FIVE TIMES MORE work!

This kind of software is supposed to "increase productivity" and make the process of translation easier and more efficient. THAT is not my experience. It makes the translation work more difficult, time-consuming and labor intensive and thus DEFINITELY REDUCES productivity.

Maybe you need very special documents to begin with?
For example during translation of a Japanese patent (in particular the claims) the order in which the individual portions of the sentence are arranged in Japanese differs significantly from for example English or German. So, the "alignment" of segments like "A-A", "B-B", "C-C" etc. does not work. It needs to be "A-C", "B-B", "A-C" or something like that.
I could not yet find out, if the (any) CAT tool would be able to handle that in any labor-easing form ...

But, as I already said, I am probably must too stupid.

Words

　本職は鍼灸師だが、本業が翻訳であるから、今回翻訳者の立場から発言する。　世界中のどの分野でも同じだが、その分野の達人にしか通じない用語がある。その用語を部外者に理解してもらう必要も時々発生する。例えば患者にムンテラ（日本人が作ったドイツ語！）の際病状を解説する時。しかし、使われている言葉に関して双方が類似の意味合いで理解しない限り会話（話が会うこと）が成り立たない。コミュニケーションを取ろうとしている人の自然条件（年齢、教育、社会的背景など）の相異が大きければ大きいほど通話(話が通る)は困難になる。
　自然な違いに加えて人間は独自の相異も作り上げる：最近何でも「科学的根拠」が必要だが、その根拠を見出すために、測定しようとする項目を定義し、測定を可能にする環境も整える。妨げる物は出来る限り除外する。これはframe of referenceと言い、カメラのファインダーと同様見えるものを限定する。
　だが、最初から見る（理解する）ものを限定する上、私のよう他文化圏の人間も居れば以心伝心の世界は夢の彼方まで遠ざかる｡アジアに於いて東西の医学では同一漢字を使うが、それぞれの分野で概念が丸きり異なる。患者の心臓が悪いと説明しなければならない時（例えば、broken heart＝失恋のための心気虚）、東西の概念を分けて理解してもらうのも一苦労だ。その患者が外国人であれば尚更に困難が増してしまう。
　日本に於いて1500年にも及ぶ東洋医学歴史中に生み出された発想が決して少なくないが、世界に殆ど知られていない。それは表現の曖昧さと「手書きあれども文書きなし」で表される文章力の貧弱さによるかもしれない（私は言える立場ではない！）。日本は東洋医学に関して世界に向けて大事な情報発信地であるべきだと思うから、情報を伝達する「言葉」をもう一度考え直す時期が迫って来たかもしれない。