No. It is wrong. Usually one omits a intercept term in a regression. In this case 回归系数 is useless.

这个很正常啊
回归系数大小说明不了什么问题
其大小取决于解释变量与被解释变量数量级的差异，由量纲决定

"其大小取决于解释变量与被解释变量数量级的差异，由量纲决定" is also wrong.

The 解释变量数量级 increasing 10 times, the corresponding coefficient will be reducing 10 times also.

It depends on what are you predicting. Your coefficients can be as big as thousands. Such as you study the relationship between the house prices and the number of bed rooms, your coefficient  can be several thousands.

非常感谢大家，不知道有没有“大系数”的paper或哪本书里的例子让我知道一下呢？

Thanks for your timely replys.