Almost 2k experiments have a blank value for technology

### Context

```
data_refinery=> select count(*) from experiments where technology ='';
count
-------
1931
```

An example is GSE43114.

### Problem or idea

These should not have a blank value for technology. I resurveyed GSE43114 locally and it had it's technology field properly populated.

### Solution or next step

We could maybe make a management command to fix this metadata, but I wonder what else went wrong with them. We should look into this further and see what the cost of rerunning all these experiments might be.

Start by reproducing the reported query for blank technology values and examining GSE43114, which the issue identifies as an example. No files or tests are named; determine what caused the missing metadata, whether a management command is appropriate, and estimate the cost of rerunning the affected experiments.

Written by the indexing model from the issue text.